1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.151 2005/06/10 13:39:52 tom Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
10 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
12 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
13 command itself ended in error, the underlying message about the transport
14 filter timeout was being overwritten with the pipe command error. Now the
15 underlying error message should be appended to the second error message.
17 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
19 PH/02 Reduce the amount of output that "make" produces by default. Full output
20 can still be requested.
22 PH/03 The warning log line about a condition test deferring for a "warn" verb
23 was being output only once per connection, rather than after each
24 occurrence (because it was using the same function as for successful
25 "warn" verbs). This seems wrong, so I have changed it.
27 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
28 it should not have, which might have caused a crash in the right
29 circumstances, but probably never did.
31 PH/04 Installed a modified version of Tony Finch's patch to make submission
32 mode fix the return path as well as the Sender: header line, and to
33 add a /name= option so that you can make the user's friendly name appear
36 TF/03 Added the control = fakedefer ACL modifier.
38 TF/04 Added the ratelimit ACL condition. See NewStuff for details.
40 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
42 TK/03 Merged latest SRS patch from Miles Wilton.
44 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
45 with the definition in sysexits.h (which is #included earlier).
46 Fortunately, Exim does not actually use EX_OK. The code used to try to
47 preserve the sysexits.h value, by assumimg that macro definitions were
48 scanned for macro replacements. I have been disabused of this notion,
49 so now the code just undefines EX_OK before #including unistd.h.
51 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
52 in the smtp transport. When a block could not be written in a single
53 write() function, the timeout was being re-applied to each part-write.
54 This seems wrong - if the receiver was accepting one byte at a time it
55 would take for ever. The timeout is now adjusted when this happens. It
56 doesn't have to be particularly precise.
58 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
59 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
61 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
62 without a domain if the check to <postmaster@domain> fails.
64 SC/01 Eximstats: added -xls and the ability to specify output files
65 (patch written by Frank Heydlauf).
67 SC/02 Eximstats: use FileHandles for outputing results.
69 SC/03 Eximstats: allow any combination of xls, txt, and html output.
71 SC/04 Eximstats: fixed display of large numbers with -nvr option
73 SC/05 Eximstats: fixed merging of reports with empty tables.
75 SC/06 Eximstats: added the -include_original_destination flag
77 SC/07 Eximstats: removed tabs and trailing whitespace.
79 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
81 TK/06 MBOX spool code: Add real "From " MBOX separator line
82 so the .eml file is really in mbox format (even though
83 most programs do not really care). Patch from Alex Miller.
85 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
86 The latter is generated from $received_to and is only set if the
87 message has one envelope recipient. SA can use these headers,
88 obviously out-of-the-box. Patch from Alex Miller.
90 PH/08 The ${def test on a variable was returning false if the variable's
91 value was "0", contrary to what the specification has always said!
92 The result should be true unless the variable is empty.
94 PH/09 The syntax error of a character other than { following "${if
95 def:variable_name" (after optional whitespace) was not being diagnosed.
96 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
97 accidental colon was present, for example, could give incorrect results.
99 PH/10 Tidied the code in a number of places where the st_size field of a stat()
100 result is used (not including appendfile, where other changes are about
103 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
104 This involved changing a lot of size variables from int to off_t. It
105 should work with maildirs and everything.
107 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
108 spamd dying while we are connected to it.
114 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
115 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
117 TK/02 Fix ACL "control" statement not being available in MIME ACL.
119 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
121 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
122 to test Sieve filters that use "vacation".
124 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
125 that changes the way the GnuTLS parameters are stored in the cache file.
126 The new format can be generated externally. For backward compatibility,
127 if the data in the cache doesn't make sense, Exim assumes it has read an
128 old-format file, and it generates new data and writes a new file. This
129 means that you can't go back to an older release without removing the
132 PH/03 A redirect router that has both "unseen" and "one_time" set does not
133 work if there are any delivery delays because "one_time" forces the
134 parent to be marked "delivered", so its unseen clone is never tried
135 again. For this reason, Exim now forbids the simultaneous setting of
138 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
139 redirected to themselves ("homonym" addresses). Read the long ChangeLog
140 entry if you want to know the details. The fix, however, neglected to
141 consider the case when local delivery batching is involved. The test for
142 "previously delivered" was not happening when checking to see if an
143 address could be batched with a previous (undelivered) one; under
144 certain circumstances this could lead to multiple deliveries to the same
147 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
148 in its include files, and this causes problems building Exim.
150 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
151 header_syntax) but Exim was just ignoring anything given after a slash.
152 In particular, this caused confusion with an attempt to use "verify =
153 reverse_host_lookup/defer_ok". An error is now given when options are
154 supplied for verify items that do not have them. (Maybe reverse_host_
155 lookup should have a defer_ok option, but that's a different point.)
157 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
158 defined by RFC 821) to 2048, because there were problems with some AUTH
159 commands, and RFC 1869 says the size should be increased for extended
160 SMTP commands that take arguments.
162 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
165 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
166 "unknown" error; now it says that the functionality isn't in the binary.
168 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
169 an address' error message when a string expansion fails (syntax or
170 whatever). Otherwise the password may appear in the log. Following change
171 PH/42 below, there is no longer a chance of it appearing in a bounce
174 PH/11 Installed exipick version 20050225.0 from John Jetmore.
176 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
177 of its addresses was ever tried. (Bugzilla bug #2.)
179 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
180 the result incorrectly in the debug output. (It correctly added a newline
181 to what was transported.)
183 TF/01 Added $received_time.
185 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
186 commented out examples of how to interface to a virus scanner and to
187 SpamAssassin. Also added commented examples of av_scanner and
188 spamd_address settings.
190 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
191 and controls are allowed in which ACLs. There were a couple of minor
192 errors. Some of the entries in the conditions table (which is a table of
193 where they are NOT allowed) were getting very unwieldy; rewrote them as a
194 negation of where the condition IS allowed.
196 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
198 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
199 header file does not have a version number, so I've had to invent a new
200 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
201 API. The code is untested by me (my Linux distribution still has 0.3.2 of
202 radiusclient), but it was contributed by a Radius user.
204 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
205 files or directories, for OpenSSL.
207 PH/19 When an Exim process that is running as root has to create an Exim log
208 file, it does so in a subprocess that runs as exim:exim so as to get the
209 ownership right at creation (otherwise, other Exim processes might see
210 the file with the wrong ownership). There was no test for failure of this
211 fork() call, which would lead to the process getting stuck as it waited
212 for a non-existent subprocess. Forks do occasionally fail when resources
213 run out. I reviewed all the other calls to fork(); they all seem to check
216 PH/20 When checking for unexpected SMTP input at connect time (before writing
217 the banner), Exim was not dealing correctly with a non-positive return
218 from the read() function. If the client had disconnected by this time,
219 the result was a log entry for a synchronization error with an empty
220 string after "input=" when read() returned zero. If read() returned -1
221 (an event I could not check), uninitialized data bytes were printed.
222 There were reports of junk text (parts of files, etc) appearing after
225 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
227 PH/22 Added support for macro redefinition, and (re)definition in between
228 driver and ACL definitions.
230 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
231 forgetting to use the resulting value; it was using the unexpanded value.
233 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
234 hadn't been configured. The fix is from Juergen Kreileder, who
235 understands it better than I do:
237 "Here's what I see happening with three configured cyrus_sasl
238 authenticators configured (plain, login, cram-md5):
240 On startup auth_cyrus_sasl_init() gets called for each of these.
241 This means three calls to sasl_listmech() without a specified mech_list.
242 => SASL tests which mechs of all available mechs actually work
243 => three warnings about OTP not working
244 => the returned list contains: plain, login, cram-md5, digest-md5, ...
246 With the patch, sasl_listmech() also gets called three times. But now
247 SASL's mech_list option is set to the server_mech specified in the the
248 authenticator. Or in other words, the answer from sasl_listmech()
249 gets limited to just the mech you're testing for (which is different
251 => the return list contains just 'plain' or 'login', 'cram-md5' or
252 nothing depending on the value of ob->server_mech.
254 I've just tested the patch: Authentication still works fine,
255 unavailable mechs specified in the exim configuration are still
256 caught, and the auth.log warnings about OTP are gone."
258 PH/25 When debugging is enabled, the contents of the command line are added
259 to the debugging output, even when log_selector=+arguments is not
262 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
263 answer is "GNU", and only if the return is "GNU/something" is the answer
266 PH/27 $acl_verify_message is now set immediately after the failure of a
267 verification in an ACL, and so is available in subsequent modifiers. In
268 particular, the message can be preserved by coding like this:
270 warn !verify = sender
271 set acl_m0 = $acl_verify_message
273 Previously, $acl_verify_message was set only while expanding "message"
274 and "log_message" when a very denied access.
276 PH/28 Modified OS/os.c-Linux with
278 -#ifndef OS_LOAD_AVERAGE
279 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
281 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
282 nomenclature these days.)
284 PH/29 Installed patch from the Sieve maintainer that adds the options
285 sieve_useraddress and sieve_subaddress to the redirect router.
287 PH/30 In these circumstances:
288 . Two addresses routed to the same list of hosts;
289 . First host does not offer TLS;
290 . First host accepts first address;
291 . First host gives temporary error to second address;
292 . Second host offers TLS and a TLS session is established;
293 . Second host accepts second address.
294 Exim incorrectly logged both deliveries with the TLS parameters (cipher
295 and peerdn, if requested) that were in fact used only for the second
298 PH/31 When doing a callout as part of verifying an address, Exim was not paying
299 attention to any local part prefix or suffix that was matched by the
300 router that accepted the address. It now behaves in the same way as it
301 does for delivery: the affixes are removed from the local part unless
302 rcpt_include_affixes is set on the transport.
304 PH/32 Add the sender address, as F=<...>, to the log line when logging a
305 timeout during the DATA phase of an incoming message.
307 PH/33 Sieve envelope tests were broken for match types other than :is. I have
308 applied a patch sanctioned by the Sieve maintainer.
310 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
311 the uid or gid is negative. A case of a negative gid caused this to be
312 noticed. The fix allows for either to be negative.
314 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
315 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
318 PH/36 The change PH/12 above was broken. Fixed it.
320 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
321 the grounds that routing the same address twice would always produce the
322 same answer. This might have been true once, but it is certainly no
323 longer true now. Routing a child address may depend on the previous
324 routing that produced that child. Some complicated redirection strategies
325 went wrong when messages had multiple recipients, and made Exim's
326 behaviour dependent on the order in which the addresses were given.
328 I have moved the duplicate checking until after the routing is complete.
329 Exim scans the addresses that are assigned to local and remote
330 transports, and removes any duplicates. This means that more work will be
331 done, as duplicates will always all be routed, but duplicates are
332 presumably rare, so I don't expect this is of any significance.
334 For deliveries to pipes, files, and autoreplies, the duplicate checking
335 still happens during the routing process, since they are not going to be
338 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
339 It corrects a timeout issue with spamd. This is Ian's comment: "The
340 background is that sometimes spamd either never reads data from a
341 connection it has accepted, or it never writes response data. The exiscan
342 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
343 blindly assumes that writes won't block so it may never time out."
345 PH/39 Allow G after quota size as well as K and M.
347 PH/40 The value set for $authenticated_id in an authenticator may not contain
348 binary zeroes or newlines because the value is written to log lines and
349 to spool files. There was no check on this. Now the value is run through
350 the string_printing() function so that such characters are converted to
351 printable escape sequences.
353 PH/41 $message_linecount is a new variable that contains the total number of
354 lines in the message. Compare $body_linecount, which is the count for the
357 PH/42 Exim no longer gives details of delivery errors for specific addresses in
358 bounce and delay warning messages, except in certain special cases, which
361 (a) An SMTP error message from a remote host;
362 (b) A message specified in a :fail: redirection;
363 (c) A message specified in a "fail" command in a system filter;
364 (d) A message specified in a FAIL return from the queryprogram router;
365 (e) A message specified by the cannot_route_message router option.
367 In these cases only, Exim does include the error details in bounce and
368 warning messages. There are also a few cases where bland messages such
369 as "unrouteable address" or "local delivery error" are given.
371 PH/43 $value is now also set for the "else" part of a ${run expansion.
373 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
374 being worked on, but at least Exim now implements the latest version to
377 PH/45 In a pipe transport, although a timeout while waiting for the pipe
378 process to complete was treated as a delivery failure, a timeout while
379 writing the message to the pipe was logged, but erroneously treated as a
380 successful delivery. Such timeouts include transport filter timeouts. For
381 consistency with the overall process timeout, these timeouts are now
382 treated as errors, giving rise to delivery failures by default. However,
383 there is now a new Boolean option for the pipe transport called
384 timeout_defer, which, if set TRUE, converts the failures into defers for
385 both kinds of timeout. A transport filter timeout is now identified in
388 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
389 systems where "make" and "gmake" are different, calling "gmake" at top
390 level broke things. I've arranged for the value of $(MAKE) to be passed
391 from the Makefile to this script so that it can call the same version of
395 A note about Exim versions 4.44 and 4.50
396 ----------------------------------------
398 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
399 changes of various kinds. As a consequence, a big documentation update was
400 needed. This delayed the release for rather longer than seemed good, especially
401 in the light of a couple of (minor) security issues. Therefore, the changes
402 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
403 release. So 4.44 and 4.50 are in effect two different branches that both start
406 I have left the 4.50 change log unchanged; it contains all the changes since
407 4.43. The change log for 4.44 is below; many of its items are identical to
408 those for 4.50. This seems to be the most sensible way to preserve the
409 historical information.
415 1. Minor wording change to the doc/README.SIEVE file.
417 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
418 computation of the current number of files was incorrect.
420 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
421 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
424 4. Give more explanation in the error message when the command for a transport
425 filter fails to execute.
427 5. There are several places where Exim runs a non-Exim command in a
428 subprocess. The SIGUSR1 signal should be disabled for these processes. This
429 was being done only for the command run by the queryprogram router. It is
430 now done for all such subprocesses. The other cases are: ${run, transport
431 filters, and the commands run by the lmtp and pipe transports.
433 6. Added CONFIGURE_GROUP build-time option.
435 7. Some older OS have a limit of 256 on the maximum number of file
436 descriptors. Exim was using setrlimit() to set 1000 as a large value
437 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
438 systems. I've change it so that if it can't get 1000, it tries for 256.
440 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
441 was an oversight, and furthermore, ever since the addition of extra
442 controls (e.g. 4.43/32), the checks on when to allow different forms of
443 "control" were broken. There should now be diagnostics for all cases when a
444 control that does not make sense is encountered.
446 9. Added the /retain_sender option to "control=submission".
448 10. $recipients is now available in the predata ACL (oversight).
450 11. Tidy the search cache before the fork to do a delivery from a message
451 received from the command line. Otherwise the child will trigger a lookup
452 failure and thereby defer the delivery if it tries to use (for example) a
453 cached ldap connection that the parent has called unbind on.
455 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
456 of $address_data from the recipient verification was clobbered by the
459 13. The value of address_data from a sender verification is now available in
460 $sender_address_data in subsequent conditions in the ACL statement.
462 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
464 15. Added a new option "connect=<time>" to callout options, to set a different
467 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
468 was its contents. (It was OK if the option was not defined at all.)
470 17. A "Completed" log line is now written for messages that are removed from
471 the spool by the -Mrm option.
473 18. New variables $sender_verify_failure and $recipient_verify_failure contain
474 information about exactly what failed.
476 19. Added -dd to debug only the daemon process.
478 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
479 handles timeouts, both on the server side and network timeouts. Renamed the
480 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
482 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
483 It is now set to "smtps".
485 22. $host_address is now set to the target address during the checking of
488 23. When checking ignore_target_hosts for an ipliteral router, no host name was
489 being passed; this would have caused $sender_host_name to have been used if
490 matching the list had actually called for a host name (not very likely,
491 since this list is usually IP addresses). A host name is now passed as
494 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
495 code that specifies a non-restarting handler (typically sigaction() in
496 modern systems) in an attempt to fix a rare and obscure crash bug.
498 25. Narrowed the window for a race in the daemon that could cause it to ignore
499 SIGCHLD signals. This is not a major problem, because they are used only to
500 wake it up if nothing else does.
502 26. A malformed maildirsize file could cause Exim to calculate negative values
503 for the mailbox size or file count. Odd effects could occur as a result.
504 The maildirsize information is now recalculated if the size or filecount
507 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
508 support for a long time. Removed HAVE_SYS_VFS_H.
510 28. Installed the latest version of exipick from John Jetmore.
512 29. In an address list, if the pattern was not a regular expression, an empty
513 subject address (from a bounce message) matched only if the pattern was an
514 empty string. Non-empty patterns were not even tested. This was the wrong
515 because it is perfectly reasonable to use an empty address as part of a
516 database query. An empty address is now tested by patterns that are
517 lookups. However, all the other forms of pattern expect the subject to
518 contain a local part and a domain, and therefore, for them, an empty
519 address still always fails if the pattern is not itself empty.
521 30. Exim went into a mad DNS loop when attempting to do a callout where the
522 host was specified on an smtp transport, and looking it up yielded more
525 31. Re-factored the code for checking spool and log partition space into a
526 function that finds that data and another that does the check. The former
527 is then used to implement four new variables: $spool_space, $log_space,
528 $spool_inodes, and $log_inodes.
530 32. The RFC2047 encoding function was originally intended for short strings
531 such as real names; it was not keeping to the 75-character limit for
532 encoded words that the RFC imposes. It now respects the limit, and
533 generates multiple encoded words if necessary. To be on the safe side, I
534 have increased the buffer size for the ${rfc2047: expansion operator from
537 33. It is now permitted to omit both strings after an "if" condition; if the
538 condition is true, the result is "true". As before, when the second string
539 is omitted, a false condition yields an empty string. This makes it less
540 cumbersome to write custom ACL and router conditions.
542 34. Failure to deliver a bounce message always caused it to be frozen, even if
543 there was an errors_to setting on the router. The errors_to setting is now
546 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
547 canonical form (fully expanded) before being placed in
548 $sender_host_address.
550 36. The table in the code that translates DNS record types into text (T_A to
551 "A" for instance) was missing entries for NS and CNAME. It is just possible
552 that this could have caused confusion if both these types were looked up
553 for the same domain, because the text type is used as part of Exim's
554 per-process caching. But the chance of anyone hitting this buglet seems
557 37. The dnsdb lookup has been extended in a number of ways.
559 (1) There is a new type, "zns", which walks up the domain tree until it
560 finds some nameserver records. It should be used with care.
562 (2) There is a new type, "mxh", which is like "mx" except that it returns
563 just the host names, not the priorities.
565 (3) It is now possible to give a list of domains (or IP addresses) to be
566 looked up. The behaviour when one of the lookups defers can be
567 controlled by a keyword.
569 (4) It is now possible to specify the separator character for use when
570 multiple records are returned.
572 38. The dnslists ACL condition has been extended: it is now possible to supply
573 a list of IP addresses and/or domains to be looked up in a particular DNS
576 39. Added log_selector=+queue_time_overall.
578 40. When running the queue in the test harness, wait just a tad after forking a
579 delivery process, to get repeatability of debugging output.
581 41. Include certificate and key file names in error message when GnuTLS fails
582 to set them up, because the GnuTLS error message doesn't include the name
583 of the failing file when there is a problem reading it.
585 42. Allow both -bf and -bF in the same test run.
587 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
589 44. The "Exiscan patch" is now merged into the mainline Exim source.
591 45. Sometimes the final signoff response after QUIT could fail to get
592 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
593 < 0 before doing a fflush(). This bug looks as though it goes back to the
594 introduction of TLS in release 3.20, but "sometimes" must have been rare
595 because the tests only now provoked it.
597 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
598 (this can affect the format of dates).
600 47. exim_tidydb, when checking for the continued existence of a message for
601 which it has found a message-specific retry record, was not finding
602 messages that were in split spool directories. Consequently, it was
603 deleting retry records that should have stayed in existence.
605 48. Steve fixed some bugs in eximstats.
607 49. The SPA authentication driver was not abandoning authentication and moving
608 on to the next authenticator when an expansion was forced to fail,
609 contradicting the general specification for all authenticators. Instead it
610 was generating a temporary error. It now behaves as specified.
612 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
613 (the order specifies the preference for clients). The order is now AES256,
614 AES128, 3DES, ARCFOUR128.
616 51. Small patch to Sieve code - explicitly set From: when generating an
619 52. Exim crashed if a remote delivery caused a very long error message to be
620 recorded - for instance if somebody sent an entire SpamAssassin report back
621 as a large number of 550 error lines. This bug was coincidentally fixed by
622 increasing the size of one of Exim's internal buffers (big_buffer) that
623 happened as part of the Exiscan merge. However, to be on the safe side, I
624 have made the code more robust (and fixed the comments that describe what
627 53. Now that there can be additional text after "Completed" in log lines (if
628 the queue_time_overall log selector is set), a one-byte patch to exigrep
629 was needed to allow it to recognize "Completed" as not the last thing in
632 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
633 patch that reportedly fixes this has been added. I am not expert enough to
634 create a test for it. This is what the patch creator wrote:
636 "I found a little strange behaviour of ldap code when working with
637 Windows 2003 AD Domain, where users was placed in more than one
638 Organization Units. When I tried to give exim partial DN, the exit code
639 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
640 But simultaneously result of request was absolutely normal ldap result,
641 so I produce this patch..."
643 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
644 REFERENCE, so I have modified the code to exclude the patch when that macro
647 55. Some experimental protocols are using DNS PTR records for new purposes. The
648 keys for these records are domain names, not reversed IP addresses. The
649 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
650 leaves it alone. Component reversal etc. now happens only for IP addresses.
653 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
655 57. Double the size of the debug message buffer (to 2048) so that more of very
656 long debug lines gets shown.
658 58. The exicyclog utility now does better if the number of log files to keep
659 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
661 59. Two changes related to the smtp_active_hostname option:
663 (1) $smtp_active_hostname is now available as a variable.
664 (2) The default for smtp_banner uses $smtp_active_hostname instead
665 of $primary_hostname.
667 60. The host_aton() function is supposed to be passed a string that is known
668 to be a valid IP address. However, in the case of IPv6 addresses, it was
669 not checking this. This is a hostage to fortune. Exim now panics and dies
670 if the condition is not met. A case was found where this could be provoked
671 from a dnsdb PTR lookup with an IPv6 address that had more than 8
672 components; fortuitously, this particular loophole had already been fixed
673 by change 4.50/55 above.
675 If there are any other similar loopholes, the new check in host_aton()
676 itself should stop them being exploited. The report I received stated that
677 data on the command line could provoke the exploit when Exim was running as
678 exim, but did not say which command line option was involved. All I could
679 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
683 61. There was a buffer overflow vulnerability in the SPA authentication code
684 (which came originally from the Samba project). I have added a test to the
685 spa_base64_to_bits() function which I hope fixes it.
688 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
689 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
691 63. The daemon start-up calls getloadavg() while still root for those OS that
692 need the first call to be done as root, but it missed one case: when
693 deliver_queue_load_max is set with deliver_drop_privilege. This is
694 necessary for the benefit of the queue runner, because there is no re-exec
695 when deliver_drop_privilege is set.
697 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
700 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
701 was in use, was not putting the data itself into the right store pool;
702 consequently, it could be overwritten for a subsequent message in the same
703 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
706 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
708 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
709 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
710 the function were treating the return as a boolean value, which happened to
711 work because 0=false and not-0=true, but is not correct code.
713 68. The host_aton() function was not handling scoped IPv6 addresses (those
714 with, for example, "%eth0" on the end) correctly.
716 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
717 negated items (that is, ~something) in unsigned ints. Some compilers
718 apparently mutter when there is no cast.
720 70. If an address verification called from an ACL failed, and did not produce a
721 user-specific message (i.e. there was only a "system" message), nothing was
722 put in $acl_verify_message. In this situation, it now puts the system
725 71. Change 4.23/11 added synchronization checking at the start of an SMTP
726 session; change 4.31/43 added the unwanted input to the log line - except
727 that it did not do this in the start of session case. It now does.
729 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
730 This is wrong and can cause the other end to generate a synchronization
731 error if it is another Exim or anything else that does the synchronization
732 check. A QUIT command is no longer sent after a timeout.
734 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
737 74. The defer_ok option of callout verification was not working if it was used
738 when verifying addresses in header lines, that is, for this case:
740 verify = header_sender/callout=defer_ok
742 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
743 those file descriptors could be used for SMTP connections. If anything
744 wrote to stderr (the example that came up was "warn" in embedded Perl), it
745 could be sent to the SMTP client, causing chaos. The daemon now opens
746 stdin, stdout, and stderr to /dev/null when it puts itself into the
749 76. Arrange for output from Perl's "warn" command to be written to Exim's main
750 log by default. The user can override this with suitable Perl magic.
752 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
753 the log message when discard triggers, was not working for the DATA ACL or
754 for the non-SMTP ACL.
756 78. Error message wording change in sieve.c.
758 79. If smtp_accept_max_per_host was set, the number of connections could be
759 restricted to fewer than expected, because the daemon was trying to set up
760 a new connection before checking whether the processes handling previous
761 connections had finished. The check for completed processes is now done
762 earlier. On busy systems, this bug wouldn't be noticed because something
763 else would have woken the daemon, and it would have reaped the completed
766 80. If a message was submitted locally by a user whose login name contained one
767 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
768 It caused a spool format error. I have fixed the spool reading code. A
769 related problem was that the "from" clause in the Received: line became
770 illegal because of the space(s). It is now covered by ${quote_local_part.
772 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
775 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
777 83. Patch from Sieve maintainer to fix unterminated string problem in
780 84. Some minor changes to the Linux configuration files to help with other
781 OS variants using glibc.
783 85. One more patch for Sieve to update vacation handling to latest spec.
786 ----------------------------------------------------
787 See the note above about the 4.44 and 4.50 releases.
788 ----------------------------------------------------
794 1. Change 4.43/35 introduced a bug that caused file counts to be
795 incorrectly computed when quota_filecount was set in an appendfile
798 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
799 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
802 3. Give more explanation in the error message when the command for a transport
803 filter fails to execute.
805 4. There are several places where Exim runs a non-Exim command in a
806 subprocess. The SIGUSR1 signal should be disabled for these processes. This
807 was being done only for the command run by the queryprogram router. It is
808 now done for all such subprocesses. The other cases are: ${run, transport
809 filters, and the commands run by the lmtp and pipe transports.
811 5. Some older OS have a limit of 256 on the maximum number of file
812 descriptors. Exim was using setrlimit() to set 1000 as a large value
813 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
814 systems. I've change it so that if it can't get 1000, it tries for 256.
816 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
817 was an oversight, and furthermore, ever since the addition of extra
818 controls (e.g. 4.43/32), the checks on when to allow different forms of
819 "control" were broken. There should now be diagnostics for all cases when a
820 control that does not make sense is encountered.
822 7. $recipients is now available in the predata ACL (oversight).
824 8. Tidy the search cache before the fork to do a delivery from a message
825 received from the command line. Otherwise the child will trigger a lookup
826 failure and thereby defer the delivery if it tries to use (for example) a
827 cached ldap connection that the parent has called unbind on.
829 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
830 of $address_data from the recipient verification was clobbered by the
833 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
834 was its contents. (It was OK if the option was not defined at all.)
836 11. A "Completed" log line is now written for messages that are removed from
837 the spool by the -Mrm option.
839 12. $host_address is now set to the target address during the checking of
842 13. When checking ignore_target_hosts for an ipliteral router, no host name was
843 being passed; this would have caused $sender_host_name to have been used if
844 matching the list had actually called for a host name (not very likely,
845 since this list is usually IP addresses). A host name is now passed as
848 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
849 code that specifies a non-restarting handler (typically sigaction() in
850 modern systems) in an attempt to fix a rare and obscure crash bug.
852 15. Narrowed the window for a race in the daemon that could cause it to ignore
853 SIGCHLD signals. This is not a major problem, because they are used only to
854 wake it up if nothing else does.
856 16. A malformed maildirsize file could cause Exim to calculate negative values
857 for the mailbox size or file count. Odd effects could occur as a result.
858 The maildirsize information is now recalculated if the size or filecount
861 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
862 support for a long time. Removed HAVE_SYS_VFS_H.
864 18. Updated exipick to current release from John Jetmore.
866 19. Allow an empty sender to be matched against a lookup in an address list.
867 Previously the only cases considered were a regular expression, or an
870 20. Exim went into a mad DNS lookup loop when doing a callout where the
871 host was specified on the transport, if the DNS lookup yielded more than
874 21. The RFC2047 encoding function was originally intended for short strings
875 such as real names; it was not keeping to the 75-character limit for
876 encoded words that the RFC imposes. It now respects the limit, and
877 generates multiple encoded words if necessary. To be on the safe side, I
878 have increased the buffer size for the ${rfc2047: expansion operator from
881 22. Failure to deliver a bounce message always caused it to be frozen, even if
882 there was an errors_to setting on the router. The errors_to setting is now
885 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
886 canonical form (fully expanded) before being placed in
887 $sender_host_address.
889 24. Updated eximstats to version 1.33
891 25. Include certificate and key file names in error message when GnuTLS fails
892 to set them up, because the GnuTLS error message doesn't include the name
893 of the failing file when there is a problem reading it.
895 26. Expand error message when OpenSSL has problems setting up cert/key files.
898 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
899 (this can affect the format of dates).
901 28. exim_tidydb, when checking for the continued existence of a message for
902 which it has found a message-specific retry record, was not finding
903 messages that were in split spool directories. Consequently, it was
904 deleting retry records that should have stayed in existence.
906 29. eximstats updated to version 1.35
907 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
908 1.35 - bugfix such that pie charts by volume are generated correctly
910 30. The SPA authentication driver was not abandoning authentication and moving
911 on to the next authenticator when an expansion was forced to fail,
912 contradicting the general specification for all authenticators. Instead it
913 was generating a temporary error. It now behaves as specified.
915 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
916 (the order specifies the preference for clients). The order is now AES256,
917 AES128, 3DES, ARCFOUR128.
919 31. Small patch to Sieve code - explicitly set From: when generating an
922 32. Exim crashed if a remote delivery caused a very long error message to be
923 recorded - for instance if somebody sent an entire SpamAssassin report back
924 as a large number of 550 error lines. This bug was coincidentally fixed by
925 increasing the size of one of Exim's internal buffers (big_buffer) that
926 happened as part of the Exiscan merge. However, to be on the safe side, I
927 have made the code more robust (and fixed the comments that describe what
930 33. Some experimental protocols are using DNS PTR records for new purposes. The
931 keys for these records are domain names, not reversed IP addresses. The
932 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
933 leaves it alone. Component reversal etc. now happens only for IP addresses.
936 34. The host_aton() function is supposed to be passed a string that is known
937 to be a valid IP address. However, in the case of IPv6 addresses, it was
938 not checking this. This is a hostage to fortune. Exim now panics and dies
939 if the condition is not met. A case was found where this could be provoked
940 from a dnsdb PTR lookup with an IPv6 address that had more than 8
941 components; fortuitously, this particular loophole had already been fixed
942 by change 4.50/55 or 4.44/33 above.
944 If there are any other similar loopholes, the new check in host_aton()
945 itself should stop them being exploited. The report I received stated that
946 data on the command line could provoke the exploit when Exim was running as
947 exim, but did not say which command line option was involved. All I could
948 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
952 35. There was a buffer overflow vulnerability in the SPA authentication code
953 (which came originally from the Samba project). I have added a test to the
954 spa_base64_to_bits() function which I hope fixes it.
957 36. The daemon start-up calls getloadavg() while still root for those OS that
958 need the first call to be done as root, but it missed one case: when
959 deliver_queue_load_max is set with deliver_drop_privilege. This is
960 necessary for the benefit of the queue runner, because there is no re-exec
961 when deliver_drop_privilege is set.
963 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
964 was in use, was not putting the data itself into the right store pool;
965 consequently, it could be overwritten for a subsequent message in the same
966 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
969 38. Sometimes the final signoff response after QUIT could fail to get
970 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
971 < 0 before doing a fflush(). This bug looks as though it goes back to the
972 introduction of TLS in release 3.20, but "sometimes" must have been rare
973 because the tests only now provoked it.
979 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
980 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
981 Now it returns an integer. A number of calls were still expecting a T/F
982 return. Fortuitously, in all cases, the tests worked in OK situations,
983 which is the norm. However, things would have gone wrong on any write
984 failures on the smtp file descriptor. This function is used when sending
985 messages over SMTP and also when doing verify callouts.
987 2. When Exim is called to do synchronous delivery of a locally submitted
988 message (the -odf or -odi options), it no longer closes stderr before doing
991 3. Implemented the mua_wrapper option.
993 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
995 5. Implemented the functions header_remove(), header_testname(),
996 header_add_at_position(), and receive_remove_recipient(), and exported them
999 6. If an ACL "warn" statement specified the addition of headers, Exim already
1000 inserted X-ACL-Warn: at the start if there was no header name. However, it
1001 was not making this test for the second and subsequent header lines if
1002 there were newlines in the string. This meant that an invalid header could
1003 be inserted if Exim was badly configured.
1005 7. Allow an ACL "warn" statement to add header lines at the start or after all
1006 the Received: headers, as well as at the end.
1008 8. Added the rcpt_4xx retry error code.
1010 9. Added postmaster_mailfrom=xxx to callout verification option.
1012 10. Added mailfrom=xxxx to the callout verification option, for verify=
1015 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1016 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1018 12. Inserted some casts to stop certain compilers warning when using pointer
1019 differences as field lengths or precisions in printf-type calls (mostly
1020 affecting debugging statements).
1022 13. Added optional readline() support for -be (dynamically loaded).
1024 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1025 same clock tick as a message's arrival, so that its received time was the
1026 same as the "first fail" time on the retry record, and that message
1027 remained on the queue past the ultimate address timeout, every queue runner
1028 would try a delivery (because it was past the ultimate address timeout) but
1029 after another failure, the ultimate address timeout, which should have then
1030 bounced the address, did not kick in. This was a "< instead of <=" error;
1031 in most cases the first failure would have been in the next clock tick
1032 after the received time, and all would be well.
1034 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1035 being recognized when the domain list was tested by the match_domain
1036 condition in an expansion string.
1038 16. Added the ${str2b64: operator.
1040 17. Exim was always calling setrlimit() to set a large limit for the number of
1041 processes, without checking whether the existing limit was already
1042 adequate. (It did check for the limit on file descriptors.) Furthermore,
1043 errors from getrlimit() and setrlimit() were being ignored. Now they are
1044 logged to the main and panic logs, but Exim does carry on, to try to do its
1045 job under whatever limits there are.
1047 18. Imported PCRE 5.0.
1049 19. Trivial typo in log message " temporarily refused connection" (the leading
1052 20. If the log selector return_path_on_delivery was set and an address was
1053 redirected to /dev/null, the delivery process crashed because it assumed
1054 that a return path would always be set for a "successful" delivery. In this
1055 case, the whole delivery is bypassed as an optimization, and therefore no
1058 21. Internal re-arrangement: the function for sending a challenge and reading
1059 a response while authentication was assuming a zero-terminated challenge
1060 string. It's now changed to take a pointer and a length, to allow for
1061 binary data in such strings.
1063 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1065 23. Exim was not respecting finduser_retries when seeking the login of the
1066 uid under which it was called; it was always trying 10 times. (The default
1067 setting of finduser_retries is zero.) Also, it was sleeping after the final
1068 failure, which is pointless.
1070 24. Implemented tls_on_connect_ports.
1072 25. Implemented acl_smtp_predata.
1074 26. If the domain in control=submission is set empty, Exim assumes that the
1075 authenticated id is a complete email address when it generates From: or
1076 Sender: header lines.
1078 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1079 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1080 chown and chgrp in /bin and hostname in /usr/bin.
1082 28. Exim was keeping the "process log" file open after each use, just as it
1083 does for the main log. This opens the possibility of it remaining open for
1084 long periods when the USR1 signal hits a daemon. Occasional processlog
1085 errors were reported, that could have been caused by this. Anyway, it seems
1086 much more sensible not to leave this file open at all, so that is what now
1089 29. The long-running daemon process does not normally write to the log once it
1090 has entered its main loop, and it closes the log before doing so. This is
1091 so that log files can straightforwardly be renamed and moved. However,
1092 there are a couple of unusual error situations where the daemon does write
1093 log entries, and I had neglected to close the log afterwards.
1095 30. The text of an SMTP error response that was received during a remote
1096 delivery was being truncated at 512 bytes. This is too short for some of
1097 the long messages that one sometimes sees. I've increased the limit to
1100 31. It is now possible to make retry rules that apply only when a message has a
1101 specific sender, in particular, an empty sender.
1103 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1104 it possible to be selective about when SMTP synchronization is enforced.
1106 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1108 32. Implemented hosts_connection_nolog.
1110 33. Added an ACL for QUIT.
1112 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1115 35. Added mailbox_size and mailbox_filecount to appendfile.
1117 36. Added control = no_multiline_responses to ACLs.
1119 37. There was a bug in the logic of the code that waits for the clock to tick
1120 in the case where the clock went backwards by a substantial amount such
1121 that the microsecond fraction of "now" was more than the microsecond
1122 fraction of "then" (but the whole seconds number was less).
1124 38. Added support for the libradius Radius client library this is found on
1125 FreeBSD (previously only the radiusclient library was supported).
1131 1. When certain lookups returned multiple values in the form name=value, the
1132 quoting of the values was not always being done properly. Specifically:
1133 (a) If the value started with a double quote, but contained no whitespace,
1135 (b) If the value contained whitespace other than a space character (i.e.
1136 tabs or newlines or carriage returns) it was not quoted.
1137 This fix has been applied to the mysql and pgsql lookups by writing a
1138 separate quoting function and calling it from the lookup code. The fix
1139 should probably also be applied to nisplus, ibase and oracle lookups, but
1140 since I cannot test any of those, I have not disturbed their existing code.
1142 2. A hit in the callout cache for a specific address caused a log line with no
1143 reason for rejecting RCPT. Now it says "Previous (cached) callout
1144 verification failure".
1146 3. There was an off-by-one bug in the queryprogram router. An over-long
1147 return line was truncated at 256 instead of 255 characters, thereby
1148 overflowing its buffer with the terminating zero. As well as fixing this, I
1149 have increased the buffer size to 1024 (and made a note to document this).
1151 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1152 when Exim is waiting for an SMTP response from a remote server, Exim
1153 restarts its select() call on the socket, thereby resetting its timeout.
1154 This is not a problem when such interrupts are rare. Somebody set up a cron
1155 job to run exiwhat every 2 minutes, which is less than the normal select()
1156 timeout (5 or 10 minutes). This meant that the select() timeout never
1157 kicked in because it was always reset. I have fixed this by comparing the
1158 time when an interrupt arrives with the time at the start of the first call
1159 to select(). If more time than the timeout has elapsed, the interrupt is
1160 treated as a timeout.
1162 5. Some internal re-factoring in preparation for the addition of Sieve
1163 extensions (by MH). In particular, the "personal" test is moved to a
1164 separate function, and given an option for scanning Cc: and Bcc: (which is
1165 not set for Exim filters).
1167 6. When Exim created an email address using the login of the caller as the
1168 local part (e.g. when creating a From: or Sender: header line), it was not
1169 quoting the local part when it contained special characters such as @.
1171 7. Installed new OpenBSD configuration files.
1173 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1174 try to make them clearer.
1176 9. Callout options, other than the timeout value, were being ignored when
1177 verifying sender addresses in header lines. For example, when using
1179 verify = header_sender/callout=no_cache
1181 the cache was (incorrectly) being used.
1183 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1184 only the Cygwin environment.
1186 11. When return_path_on_delivery was set as a log selector, if different remote
1187 addresses in the same message used different return paths and parallel
1188 remote delivery occurred, the wrong values would sometimes be logged.
1189 (Whenever a remote delivery process finished, the return path value from
1190 the most recently started remote delivery process was logged.)
1192 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1193 lines when AUTH and/or TLS are in use. This is the "received protocol"
1194 field. Exim used to use "asmtp" for authenticated SMTP, without any
1195 indication (in the protocol name) for TLS use. Now it follows the RFC and
1196 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1197 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1198 names appear in log lines as well as in Received: header lines.
1200 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1201 extensions, and comparison tests, and to fix some bugs.
1203 14. Changes to the "personal" filter test:
1205 (1) The test was buggy in that it was just doing the equivalent of
1206 "contains" tests on header lines. For example, if a user's address was
1207 anne@some.where, the "personal" test would incorrectly be true for
1209 To: susanne@some.where
1211 This test is now done by extracting each address from the header in turn,
1212 and checking the entire address. Other tests that are part of "personal"
1213 are now done using regular expressions (for example, to check local parts
1214 of addresses in From: header lines).
1216 (2) The list of non-personal local parts in From: addresses has been
1217 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1218 taken from the Sieve specification recommendations.
1220 (3) If the message contains any header line starting with "List-" it is
1221 treated as non-personal.
1223 (4) The test for "circular" in the Subject: header line has been removed
1224 because it now seems ill-conceived.
1226 15. Minor typos in src/EDITME comments corrected.
1228 16. Installed latest exipick from John Jetmore.
1230 17. If headers_add on a router specified a text string that was too long for
1231 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1232 of string_sprintf() is now avoided.
1234 18. $message_body_size was not set (it was always zero) when running the DATA
1235 ACL and the local_scan() function.
1237 19. For the "mail" command in an Exim filter, no default was being set for
1238 the once_repeat time, causing a random time value to be used if "once" was
1239 specified. (If the value happened to be <= 0, no repeat happened.) The
1240 default is now 0s, meaning "never repeat". The "vacation" command was OK
1241 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1242 (I found it when inspecting the code).
1244 20. There is now an overall timeout for performing a callout verification. It
1245 defaults to 4 times the callout timeout, which applies to individual SMTP
1246 commands during the callout. The overall timeout applies when there is more
1247 than one host that can be tried. The timeout is checked before trying the
1248 next host. This prevents very long delays if there are a large number of
1249 hosts and all are timing out (e.g. when the network connections are timing
1250 out). The value of the overall timeout can be changed by specifying an
1251 additional sub-option for "callout", called "maxwait". For example:
1253 verify = sender/callout=5s,maxwait=20s
1255 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1256 to the end before writing, but this should make it even safer).
1258 22. Exim was forgetting that it had advertised PIPELINING for the second and
1259 subsequent messages on an SMTP connection. It was also not resetting its
1260 memory on STARTTLS and an internal HELO.
1262 23. When Exim logs an SMTP synchronization error within a session, it now
1263 records whether PIPELINING has been advertised or not.
1265 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1266 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1267 rather than long int.
1269 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1271 26. Added the never_mail option to autoreply.
1277 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1278 crash if the getsockname() call failed; this can happen if a connection is
1279 closed very soon after it is established. The problem was simply in the
1280 order in which certain operations were done, causing Exim to try to write
1281 to the SMTP stream before it had set up the file descriptor. The bug has
1282 been fixed by making things happen in the correct order.
1288 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1289 before the connection was closed, thus losing the rejection response.
1291 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1292 some early Solaris releases, but causes trouble in current releases where
1293 socklen_t is defined.
1295 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1298 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1301 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1302 that releasing the top of it at the end releases what was used for sub-
1303 expansions (unless the block got too big). However, discard this block if
1304 the first thing is a variable or header, so that we can use its block when
1305 it is dynamic (useful for very large $message_headers, for example).
1307 6. Lookups now cache *every* query, not just the most recent. A new, separate
1308 store pool is used for this. It can be recovered when all lookup caches are
1309 flushed. Lookups now release memory at the end of their result strings.
1310 This has involved some general refactoring of the lookup sources.
1312 7. Some code has been added to the store_xxx() functions to reduce the amount
1313 of flapping under certain conditions.
1315 8. log_incoming_interface used to affect only the <= reception log lines. Now
1316 it causes the local interface and port to be added to several more SMTP log
1317 lines, for example "SMTP connection from", and rejection lines.
1319 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1321 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1323 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1324 could be overwritten at the end of the current message (or the start of a
1325 new message if it was set in a HELO ACL). The value is now preserved for
1326 the duration of the SMTP connection.
1328 12. If a transport had a headers_rewrite setting, and a matching header line
1329 contained an unqualified address, that address was qualified, even if it
1330 did not match any rewriting rules. The underlying bug was that the values
1331 of the flags that permit the existence of unqualified sender and recipient
1332 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1333 non-local messages, and by -bnq for local messages) were not being
1334 preserved with the message after it was received.
1336 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1337 "next input=" as part of the text comprising the host identity instead of
1338 the correct text. The code was using the same buffer for two different
1339 strings. However, depending on which order the printing function evaluated
1340 its arguments, the bug did not always show up. Under Linux, for example, my
1341 test suite worked just fine.
1343 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1344 doesn't work with some older versions of Perl. It has been changed to "my",
1345 which in any case is probably the better facility to use.
1347 15. A really picky compiler found some instances of statements for creating
1348 error messages that either had too many or two few arguments for the format
1351 16. The size of the buffer for calls to the DNS resolver has been increased
1352 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1353 for addresses that have a lot of PTR records. This alleviates a problem; it
1354 does not fully solve it.
1356 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1357 buffer now truncates the list and logs the incident, which is the same
1358 action as happens when Exim is looking up a host name and its aliases.
1359 Previously in this situation something unpredictable would happen;
1360 sometimes it was "internal error: store_reset failed".
1362 18. If a server dropped the connection unexpectedly when an Exim client was
1363 using GnuTLS and trying to read a response, the client delivery process
1364 crashed while trying to generate an error log message.
1366 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1367 string, for example:
1369 warn message = H1: something\nH2: something
1371 the text was added as a single header line from Exim's point of view
1372 though it ended up OK in the delivered message. However, searching for the
1373 second and subsequent header lines using $h_h2: did not work. This has been
1374 fixed. Similarly, if a system filter added multiple headers in this way,
1375 the routers could not see them.
1377 20. Expanded the error message when iplsearch is called with an invalid key to
1378 suggest using net-iplsearch in a host list.
1380 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1381 ACLs are no longer actually imposed (and a message to that effect is
1384 22. If a "gecos" field in a passwd entry contained escaped characters, in
1385 particular, if it contained a \" sequence, Exim got it wrong when building
1386 a From: or a Sender: header from that name. A second bug also caused
1387 incorrect handling when an unquoted " was present following a character
1388 that needed quoting.
1390 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1391 was not being matched caselessly.
1393 24. Arranged for all hyphens in the exim.8 source to be escaped with
1396 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1397 properly. Recipient callout cache records were still being keyed to include
1398 the sender, even when use_sender was set false. This led to far more
1399 callouts that were necessary. The sender is no longer included in the key
1400 when use_sender is false.
1402 26. Added "control = submission" modifier to ACLs.
1404 27. Added the ${base62d: operator to decode base 62 numbers.
1406 28. dnsdb lookups can now access SRV records.
1408 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1409 the configuration file.
1411 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1412 (-v) mode. This makes the output for a verbose queue run more intelligible.
1414 31. Added a use_postmaster feature to recipient callouts.
1416 32. Added the $body_zerocount variable, containing the number of binary zero
1417 bytes in the message body.
1419 33. The time of last modification of the "new" subdirectory is now used as the
1420 "mailbox time last read" when there is a quota error for a maildir
1423 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1425 35. Added +ignore_unknown as a special item in host lists.
1427 36. Code for decoding IPv6 addresses in host lists is now included, even if
1428 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1429 address was recognized as an IP address, but was then not correctly decoded
1430 into binary, causing unexpected and incorrect effects when compared with
1437 1. Very minor rewording of debugging text in manualroute to say "list of
1438 hosts" instead of "hostlist".
1440 2. If verify=header_syntax was set, and a header line with an unqualified
1441 address (no domain) and a large number of spaces between the end of the
1442 name and the colon was received, the reception process suffered a buffer
1443 overflow, and (when I tested it) crashed. This was caused by some obsolete
1444 code that should have been removed. The fix is to remove it!
1446 3. When running in the test harness, delay a bit after writing a bounce
1447 message to get a bit more predictability in the log output.
1449 4. Added a call to search_tidyup() just before forking a reception process. In
1450 theory, someone could use a lookup in the expansion of smtp_accept_max_
1451 per_host which, without the tidyup, could leave open a database connection.
1453 5. Added the variables $recipient_data and $sender_data which get set from a
1454 lookup success in an ACL "recipients" or "senders" condition, or a router
1455 "senders" option, similar to $domain_data and $local_part_data.
1457 6. Moved the writing of debug_print from before to after the "senders" test
1460 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1461 problems for message scanning, either using a data ACL, or using
1462 local_scan() because the Received: header was not generated till after they
1463 were called (in order to set the time as the time of reception completion).
1464 I have revised the way this works. The header is now generated after the
1465 body is received, but before the ACL or local_scan() are called. After they
1466 are run, the timestamp in the header is updated.
1472 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1473 before starting a queue runner without re-exec. This happened only when
1474 deliver_drop_privilege was set or when the Exim user was set to root. The
1475 effect of the bug was that timeouts during subsequent deliveries caused
1476 crashes instead of being properly handled. The handler is now left at its
1477 default (and expected) setting.
1479 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1480 message, again when deliver_drop_privilege is set or Exim is run as root.
1481 The bug described in (1) was not present in this case, but the tidying up
1482 of the other signals was missing. I have made the two cases consistent.
1484 3. The ignore_target_hosts setting on a manualroute router was being ignored
1485 for hosts that were looked up using the /MX notation.
1487 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1490 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1491 operated on the sender address. After changing the $sender_address to <>
1492 for the sender address verify, Exim was re-instated it as the original
1493 (before rewriting) address, but remembering that it had rewritten it, so it
1494 wasn't rewriting it again. This bug also had the effect of breaking the
1495 sender address verification caching when the sender address was rewritten.
1497 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1498 This has been changed so that if the ip literal address matches
1499 ignore_target_hosts, the router declines.
1501 7. Added expansion conditions match_domain, match_address, and match_local_
1502 part (NOT match_host).
1504 8. The placeholder for the Received: header didn't have a length field set.
1506 9. Added code to Exim itself and to exim_lock to test for a specific race
1507 condition that could lead to file corruption when using MBX delivery. The
1508 issue is with the lockfile that is created in /tmp. If this file is removed
1509 after a process has opened it but before that process has acquired a lock,
1510 there is the potential for a second process to recreate the file and also
1511 acquire a lock. This could lead to two Exim processes writing to the file
1512 at the same time. The added code performs the same test as UW imapd; it
1513 checks after acquiring the lock that its file descriptor still refers to
1514 the same named file.
1516 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1517 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1520 11. Added the smtp_active_hostname option. If used, this will typically be made
1521 to depend on the incoming interface address. Because $interface_address is
1522 not set up until the daemon has forked a reception process, error responses
1523 that can happen earlier (such as "too many connections") no longer contain
1526 12. If an expansion in a condition on a "warn" statement fails because a lookup
1527 defers, the "warn" statement is abandoned, and the next ACL statement is
1528 processed. Previously this caused the whole ACL to be aborted.
1530 13. Added the iplsearch lookup type.
1532 14. Added ident_timeout as a log selector.
1534 15. Added tls_certificate_verified as a log selector.
1536 16. Added a global option tls_require_ciphers (compare the smtp transport
1537 option of the same name). This controls incoming TLS connections.
1539 17. I finally figured out how to make tls_require_ciphers do a similar thing
1540 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1541 before starting the TLS session.
1543 18. Tabs are now shown as \t in -bP output.
1545 19. If the log selector return_path_on_delivery was set, Exim crashed when
1546 bouncing a message because it had too many Received: header lines.
1548 20. If two routers both had headers_remove settings, and the first one included
1549 a superfluous trailing colon, the final name in the first list and the
1550 first name in the second list were incorrectly joined into one item (with a
1551 colon in the middle).
1557 1. Added -C and -D options to the exinext utility, mainly to make it easier
1558 to include in the automated testing, but these could be helpful when
1559 multiple configurations are in use.
1561 2. The exinext utility was not formatting the output nicely when there was
1562 an alternate port involved in the retry record key, nor when there was a
1563 message id as well (for retries that were specific to a specific message
1564 and a specific host). It was also confused by IPv6 addresses, because of
1565 the additional colons they contain. I have fixed the IPv4 problem, and
1566 patched it up to do a reasonable job for IPv6.
1568 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1569 delivery, the log line now contains "pipelined" if PIPELINING was used.
1571 4. An SMTP transport process used to panic and die if the bind() call to set
1572 an explicit outgoing interface failed. This has been changed; it is now
1573 treated in the same way as a connect() failure.
1575 5. A reference to $sender_host_name in the part of a conditional expansion
1576 that was being skipped was still causing a DNS lookup. This no longer
1579 6. The def: expansion condition was not recognizing references to header lines
1580 that used bh_ and bheader_.
1582 7. Added the _cache feature to named lists.
1584 8. The code for checking quota_filecount in the appendfile transport was
1585 allowing one more file than it should have been.
1587 9. For compatibility with Sendmail, the command line option
1595 and sets the incoming protocol and host name (for trusted callers). The
1596 host name and its colon can be omitted when only the protocol is to be set.
1597 Note the Exim already has two private options, -pd and -ps, that refer to
1598 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1599 "s", but I don't think that's a major issue.
1601 10. A number of refactoring changes to the code, none of which should affect
1604 (a) The number of logging options was getting close to filling up the
1605 32-bit word that was used as a bit map. I have split them into two classes:
1606 those that are passed in the argument to log_write(), and those that are
1607 only ever tested independently outside of that function. These are now in
1608 separate 32-bit words, so there is plenty of room for expansion again.
1609 There is no change in the user interface or the logging behaviour.
1611 (b) When building, for example, log lines, the code previously used a
1612 macro that called string_cat() twice, in order to add two strings. This is
1613 not really sufficiently general. Furthermore, there was one instance where
1614 it was actually wrong because one of the argument was used twice, and in
1615 one call a function was used. (As it happened, calling the function twice
1616 did not affect the overall behaviour.) The macro has been replaced by a
1617 function that can join an arbitrary number of extra strings onto a growing
1620 (c) The code for expansion conditions now uses a table and a binary chop
1621 instead of a serial search (which was left over from when there were very
1622 few conditions). Also, it now recognizes conditions like "pam" even when
1623 the relevant support is not compiled in: a suitably worded error message is
1624 given if an attempt is made to use such a condition.
1626 11. Added ${time_interval:xxxxx}.
1628 12. A bug was causing one of the ddress fields not to be passed back correctly
1629 from remote delivery subprocesses. The field in question was not being
1630 subsequently used, so this caused to problems in practice.
1632 13. Added new log selectors queue_time and deliver_time.
1634 14. Might have fixed a bug in maildirsizefile handling that threw up
1635 "unexpected character" debug warnings, and recalculated the data
1636 unnecessarily. In any case, I expanded the warning message to give more
1639 15. Added the message "Restricted characters in address" to the statements in
1640 the default ACL that block characters like @ and % in local parts.
1642 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1643 Three changes have been made:
1645 (a) There was a serious bug; a negative response to MAIL caused the whole
1646 recipient domain to be cached as invalid, thereby blocking all messages
1647 to all local parts at the same domain, from all senders. This bug has
1648 been fixed. The domain is no longer cached after a negative response to
1649 MAIL if the sender used is not empty.
1651 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1654 (c) A new callout option, "use_sender" has been added for people who want
1655 the modified behaviour.
1661 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1664 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1665 indeed breaks things for older releases.
1667 3. Added additional logging to the case where there is a problem reading data
1668 from a filter that is running in a subprocess using a pipe, in order to
1669 try to track down a specific problem.
1671 4. Testing facility fudge: when running in the test harness and attempting
1672 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1673 getting "No route to host". Convert this to a timeout.
1675 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1678 6. Some OS don't have socklen_t but use size_t instead. This affects the
1679 fifth argument of getsockopt() amongst other things. This is now
1680 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1681 can be set for individual OS. I have set it for SunOS5, OSF1, and
1682 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1683 some earlier ones do not.
1685 7. Change 4.30/15 was not doing the test caselessly.
1687 8. The standard form for an IPv6 address literal was being rejected by address
1688 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1689 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1690 this, as well as the form without the "IPv6" on the front (but only when
1691 address literals are enabled, of course).
1693 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1695 10. Exim crashed if a message with an empty sender address specified by -f
1696 encountered a router with an errors_to setting. This could be provoked only
1697 by a command such as
1701 where an empty string was supplied; "<>" did not hit this bug.
1703 11. Installed PCRE release 4.5.
1705 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1706 remained set. It is now erased.
1708 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1709 times from message ids (which are base 36 rather than the normal 62).
1711 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1712 were being counted as actual protocol errors, and logged if the log
1713 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1714 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1715 and DATA following a set of rejected RCPTs do not count as protocol errors.
1716 In other words, Exim assumes they were pipelined, though this may not
1717 actually be the case. Of course, in all cases the client gets an
1718 appropriate error code.
1720 15. If a lookup fails in an ACL condition, a message about the failure may
1721 be available; it is used if testing the ACL cannot continue, because most
1722 such messages specify what the cause of the deferral is. However, some
1723 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1724 that caused an old message to be retained and used if a later statement
1725 caused a defer, replacing the real cause of the deferral.
1727 16. If an IP address had so many PTR records that the DNS lookup buffer
1728 was not large enough to hold them, Exim could crash while trying to process
1729 the truncated data. It now detects and logs this case.
1731 17. Further to 4.21/58, another change has been made: if (and only if) the
1732 first line of a message (the first header line) ends with CRLF, a bare LF
1733 in a subsequent header line has a space inserted after it, so as not to
1734 terminate the header.
1736 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1737 unnecessarily, used atoi() instead of strtol(), and didn't check the
1738 termination when getting file sizes from file names by regex.
1740 19. Completely re-implemented the support for maildirsize files, in the light
1741 of a number of problems with the previous contributed implementation
1742 (4.30/29). In particular:
1744 . If the quota is zero, the maildirsize file is maintained, but no quota is
1747 . If the maildir directory does not exist, it is created before any attempt
1748 to write a maildirsize file.
1750 . The quota value in the file is just a cache; if the quota is changed in
1751 the transport, the new value overrides.
1753 . A regular expression is available for excluding directories from the
1756 20. The autoreply transport checks the characters in options that define the
1757 message's headers; it allows continued headers, but it was checking with
1758 isspace() after an embedded newline instead of explicitly looking for a
1761 21. If all the "regular" hosts to which an address was routed had passed their
1762 expiry times, and had not reached their retry times, the address was
1763 bounced, even if fallback hosts were defined. Now Exim should go on to try
1766 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1767 equivalent code in the SMTP transport. Some hosts send humungous responses
1768 to HELO/EHLO, more than 1024 it seems.
1770 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1771 gives compiler warnings in some environments. I've now done it "properly",
1774 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1775 (because of problems with the built-in one) was declared to return uschar *
1776 instead of char *, causing compiler failure.
1778 25. Fixed a file descriptor leak when processing alias/forward files.
1780 26. Fixed a minor format string issue in dbfn.c.
1782 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1784 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1785 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1788 29. When a sender address is verified, it is cached, to save repeating the test
1789 when there is more than one recipient in a message. However, when the
1790 verification involves a callout, it is possible for different callout
1791 options to be set for different recipients. It is too complicated to keep
1792 track of this in the cache, so now Exim always runs a verification when a
1793 callout is required, relying on the callout cache for the optimization.
1794 The overhead is duplication of the address routing, but this should not be
1797 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
1798 to be verified with callout=postmaster, and the main callout worked but the
1799 postmaster check failed, the verification correctly failed. However, if a
1800 subsequent RCPT command asked for sender verification *without* the
1801 postmaster check, incorrect caching caused this verification also to fail,
1804 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1805 it was not caching the DNS options (qualify_single, search_parents) that
1806 were used when the lookup failed. A subsequent lookup with different
1807 options therefore always gave the same answer, though there were cases
1808 where it should not have. (Example: a "domains = !$mx_any" option on a
1809 dnslookup router: the "domains" option is always processed without any
1810 widening, but the router might have qualify_single set.) Now Exim uses the
1811 cached value only when the same options are set.
1813 32. Added John Jetmore's "exipick" utility to the distribution.
1815 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
1816 than a timeout (e.g. a certificate is required, and is not provided), an
1817 Exim server now closes the connection immediately. Previously it waited for
1818 the client to close - but if the client is SSL, it seems that they each
1819 wait for each other, leading to a delay before one of them times out.
1821 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1822 maintained 0.8.x compatibility because I don't think many are using it, and
1823 it is clearly obsolete.
1825 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1828 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1829 client certificate was expired. A simple patch fixes this, though I don't
1830 understand the full logic of why the verify callback is called multiple
1833 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1834 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1835 which causes problems with some clients (such as the Certicom SSL Plus
1836 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1837 disables the coutermeasure allowing Eudora to connect."
1839 38. Exim was not checking that a write() to a log file succeeded. This could
1840 lead to Bad Things if a log got too big, in particular if it hit a file
1841 size limit. Exim now panics and dies if it cannot write to a log file, just
1842 as it does if it cannot open a log file.
1844 39. Modified OS/Makefile-Linux so that it now contains
1846 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
1848 The two -D definitions ensure that Exim is compiled with large file
1849 support, which makes it possible to handle log files that are bigger than
1852 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
1853 instance) a domain was checked against a named list that involved a lookup,
1854 causing $domain_data to be set, then another domain was checked against the
1855 same list, then the first domain was re-checked, the value of $domain_data
1856 after the final check could be wrong. In particular, if the second check
1857 failed, it could be set empty. This bug probably also applied to
1860 41. The strip_trailing_dot option was not being applied to the address given
1861 with the -f command-line option.
1863 42. The code for reading a message's header from the spool was incrementing
1864 $received_count, but never initializing it. This meant that the value was
1865 incorrect (doubled) while delivering a message in the same process in which
1866 it was received. In the most common configuration of Exim, this never
1867 happens - a fresh exec is done - but it can happen when
1868 deliver_drop_privilege is set.
1870 43. When Exim logs an SMTP synchronization error - client data sent too soon -
1871 it now includes up to 150 characters of the unexpected data in the log
1874 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
1875 and building data strings. The size of both of these buffers was 10 000
1876 bytes - far larger than anybody would *ever* want, thought I. Needless to
1877 say, somebody hit the limit. I have increased the maximum line length to
1878 20 000 and the maximum data length of concatenated lines to 100 000. I have
1879 also fixed two bugs, because there was no checking on these buffers. Tsk,
1880 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
1881 buffer is too small.
1883 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
1884 lsearch lookups. Now it does.
1886 46. When parsing a route_list item in a manualroute router, a fixed-length
1887 buffer was used for the list of hosts. I made this 1024 bytes long,
1888 thinking that nobody would ever have a list of hosts that long. Wrong.
1889 Somebody had a whole pile of complicated expansion conditions, and the
1890 string was silently truncated, leading to an expansion error. It turns out
1891 that it is easier to change to an unlimited length (owing to other changes
1892 that have happened since this code was originally written) than to build
1893 structure for giving a limitation error. The length of the item that
1894 expands into the list of hosts is now unlimited.
1896 47. The lsearch lookup could not handle data where the length of text line was
1897 more than 4095 characters. Such lines were truncated, leading to shortened
1898 data being returned. It should now handle lines of any length.
1900 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
1901 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
1904 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
1905 replacing strings such as BIN_DIRECTORY in a source file: the replacement
1906 no longer happens in comment lines. A list of replacements is now placed
1907 at the head of all of the source files, except those whose only change is
1908 to replace PERL_COMMAND in the very first #! line.
1910 50. Replaced the slow insertion sort in queue.c, for sorting the list of
1911 messages on the queue, with a bottom-up merge sort, using code contributed
1912 by Michael Haardt. This should make operations like -bp somewhat faster on
1913 large queues. It won't affect queue runners, except when queue_run_in_order
1916 51. Installed eximstats 1.31 in the distribution.
1918 52. Added support for SRV lookups to the dnslookup router.
1920 53. If an ACL referred to $message_body or $message_body_end, the value was not
1921 reset for any messages that followed in the same SMTP session.
1923 54. The store-handling optimization for building very long strings was not
1924 differentiating between the different store pools. I don't think this
1925 actually made any difference in practice, but I've tidied it.
1927 55. While running the routers to verify a sender address, $sender_address
1928 was still set to the sender address. This is wrong, because when routing to
1929 send a bounce to the sender, it would be empty. Therefore, I have changed
1930 it so that, while verifying a sender address, $sender_address is set to <>.
1931 (There is no change to what happens when verifying a recipient address.)
1933 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
1934 target A or AAAA records (if not already returned) without resetting the
1935 qualify_single or search_parents options of the DNS resolver. These are
1936 inappropriate in this case because the targets of MX and SRV records must
1937 be FQDNs. A broken DNS record could cause trouble if it happened to have a
1938 target that, when qualified, matched something in the local domain. These
1939 two options are now turned off when doing these lookups.
1941 57. It seems that at least some releases of Reiserfs (which does not have the
1942 concept of a fixed number of inodes) returns zero and not -1 for the
1943 number of available inodes. This interacted badly with check_spool_inodes,
1944 which assumed that -1 was the "no such thing" setting. What I have done is
1945 to check that the total number of inodes is greater than zero before doing
1946 the test of how many are available.
1948 58. When a "warn" ACL statement has a log_message modifier, the message is
1949 remembered, and not repeated. This is to avoid a lot of repetition when a
1950 message has many recipients that cause the same warning to be written.
1951 Howewer, Exim was preserving the list of already written lines for an
1952 entire SMTP session, which doesn't seem right. The memory is now reset if a
1953 new message is started.
1955 59. The "rewrite" debugging flag was not showing the result of rewriting in the
1956 debugging output unless log_rewrite was also set.
1958 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
1959 of (int)(handle) when we know that handle contains (void *)(-1).
1961 61. The Exim daemon panic-logs an error return when it closes the incoming
1962 connection. However "connection reset by peer" seems to be common, and
1963 isn't really an error worthy of noting specially, so that particular error
1966 62. When Exim is trying to find all the local interfaces, it used to panic and
1967 die if the ioctl to get the interface flags failed. However, it seems that
1968 on at least one OS (Solaris 9) it is possible to have an interface that is
1969 included in the list of interfaces, but for which you get a failure error
1970 for this call. This happens when the interface is not "plumbed" into a
1971 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
1972 failure of the "get flags" call assumes that the interface is down.
1974 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
1975 makes life easier for people who are doing arithmetic on fields extracted
1976 from dates, where you often get leading zeros that should not be
1977 interpreted as octal.
1979 64. Added qualify_domain to the redirect router, to override the global
1982 65. If a pathologically long header line contained very many addresses (the
1983 report of this problem mentioned 10 000) and each of them was rewritten,
1984 Exim could use up a very large amount of memory. (It kept on making new
1985 copies of the header line as it rewrote, and never released the old ones.)
1986 At the expense of a bit more processing, the header rewriting function has
1987 been changed so that it no longer eats memory in this way.
1989 66. The generation of the Received: header has been moved from the time that a
1990 message starts to be received, to the time that it finishes. The timestamp
1991 in the Received: header should now be very close to that of the <= log
1992 line. There are two side-effects of this change:
1994 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
1995 logged header lines no longer include the local Received: line, because
1996 it has not yet been created. The same applies to a copy of the message
1997 that is returned to a non-SMTP sender when a message is rejected.
1999 (b) When a filter file is tested using -bf, no additional Received: header
2000 is added to the test message. After some thought, I decided that this
2003 This change does not affect the value of $received_for. It is still set
2004 after address rewriting, but before local_scan() is called.
2006 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2008 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2009 gave an unhelpful panic error message, and a defer error. I have managed to
2010 change this behaviour so that it now rejects any supplied certificate,
2011 which seems right, as the list of acceptable certificates is empty.
2013 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2014 gave an unhelpful defer error. I have not managed to make this reject any
2015 supplied certificates, but the error message it gives is "no certificate
2016 supplied", which is not helpful.
2018 70. exigrep's output now also includes lines that are not associated with any
2019 message, but which match the given pattern. Implemented by a patch from
2020 Martin Sluka, which also tidied up the Perl a bit.
2022 71. Recipient callout verification, like sender verification, was using <> in
2023 the MAIL FROM command. This isn't really the right thing, since the actual
2024 sender may affect whether the remote host accepts the recipient or not. I
2025 have changed it to use the actual sender in the callout; this means that
2026 the cache record is now keyed on a recipient/sender pair, not just the
2027 recipient address. There doesn't seem to be a real danger of callout loops,
2028 since a callout by the remote host to check the sender would use <>.
2029 [SEE ABOVE: changed after hitting problems.]
2031 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2032 temporary errors. However, in the case of such a code being given after
2033 the end of a data transmission (i.e. after ".") Exim was failing to write
2034 a retry record for the message. (Yes, there was some broken host that was
2035 actually sending 8xx at this point.)
2037 73. An unknown lookup type in a host list could cause Exim to panic-die when
2038 the list was checked. (An example that provoked this was putting <; in the
2039 middle of a list instead of at the start.) If this happened during a DATA
2040 ACL check, a -D file could be left lying around. This kind of configuration
2041 error no longer causes Exim to die; instead it causes a defer errror. The
2042 incident is still logged to the main and panic logs.
2044 74. Buglet left over from Exim 3 conversion. The message "too many messages
2045 in one connection" was written to the rejectlog but not the mainlog, except
2046 when address rewriting (yes!) was being logged.
2048 75. Added write_rejectlog option.
2050 76. When a system filter was run not as root (that is, when system_filter_user
2051 was set), the values of the $n variables were not being returned to the
2052 main process; thus, they were not subsequently available in the $sn
2055 77. Added +return_path_on_delivery log selector.
2057 78. A connection timeout was being treated differently from recipients deferred
2058 when testing hosts_max_try with a message that was older than the host's
2059 retry timeout. (The host should not be counted, thus allowing all hosts to
2060 be tried at least once before bouncing.) This may have been the cause of an
2061 occasionally reported bug whereby a message would remain on the queue
2062 longer than the retry timeout, but would be bounced if a delivery was
2063 forced. I say "may" because I never totally pinned down the problem;
2064 setting up timeout/retry tests is difficult. See also the next item.
2066 79. The ultimate address timeout was not being applied to errors that involved
2067 a combination of host plus message (for example, a timeout on a MAIL
2068 command). When an address resolved to a number of possible hosts, and they
2069 were not all tried for each delivery (e.g. because of hosts_max_try), a
2070 message could remain on the queue longer than the retry timeout.
2072 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2075 81. Fixed an obscure SMTP outgoing bug which required at least the following
2076 conditions: (a) there was another message waiting for the same server;
2077 (b) the server returned 5xx to all RCPT commands in the first message so
2078 that the message was not completed; (c) the server dropped the connection
2079 or gave a negative response to the RSET that Exim sends to abort the
2080 transaction. The observed case was a dropped connection after DATA that had
2081 been sent in pipelining mode. That is, the server had advertised PIPELINING
2082 but was not implementing it correctly. The effect of the bug was incorrect
2083 behaviour, such as trying another host, and this could lead to a crash.
2089 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2090 and daemon.c were passed as pointers to ints; they should have been
2091 pointers to socklen_t variables (which are typically unsigned ints).
2093 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2096 3. Fixed a really odd bug that affected only the testing scheme; patching a
2097 certain fixed string in the binary changed the value of another string that
2098 happened to be identical to the end of the original first string.
2100 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2101 name", it returns that address as the IP address. On some operating
2102 systems (e.g. Solaris), it also passes back the IP address string as the
2103 "host name". However, on others (e.g. Linux), it passes back an empty
2104 string. Exim wasn't checking for this, and was changing the host name to an
2105 empty string, assuming it had been canonicized.
2107 5. Although rare, it is permitted to have more than one PTR record for a given
2108 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2109 all the names associated with an address, because they do in Solaris.
2110 However, it seems that they do not in Linux for data that comes from the
2111 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2112 I found this out when I moved to a new Linux workstation and tried to run
2113 the Exim test suite.
2115 To get round this problem I have changed the code so that it now does its
2116 own call to the DNS to look up PTR records when searching for a host name.
2117 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2118 addresses that are only in /etc/hosts are still found.
2120 This behaviour is, however, controlled by an option called host_lookup_
2121 order, which defaults to "bydns:byaddr". If people want to use the other
2122 order, or indeed, just use one or the other means of lookup, they can
2123 specify it in this variable.
2125 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2126 some operating systems, this comes back from gethostbyaddr() as an empty
2127 string, and this is what Exim used to test for. However, it seems that in
2128 other systems, "." is yielded. Exim now tests for this case too.
2130 7. The values of check_spool_space and check_log_space are now held internally
2131 as a number of kilobytes instead of an absolute number of bytes. If a
2132 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2133 kilobyte. This means that much larger values can be stored.
2135 8. Exim monitor: an attempt to get the action menu when not actually pointing
2136 at a message produces an empty menu entitled "No message selected". This
2137 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2138 no entries in it ("Shell widget menu has zero width and/or height"). So I
2139 have added a single, blank menu entry in this case.
2141 9. Added ${quote_local_part.
2143 10. MIME decoding is now applied to the contents of Subject: header lines when
2146 11. Now that a reference to $sender_host_address automatically causes a reverse
2147 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2148 host lookup before query-style lookups in lists that might use this
2149 variable. This has therefore been abolished, and the "net-" prefix is no
2150 longer necessary for query-style lookups.
2152 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2153 have been a typo for LFLAGS, so it has been changed.
2155 13. The install script calls Exim with "-C /dev/null" in order to find the
2156 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2157 to be output. Howeve, since Exim outputs its version number before the
2158 error, it didn't break the script. It just looked ugly. I fixed this by
2159 always allowing "-C /dev/null" if the caller is root.
2161 14. Ignore overlarge ACL variable number when reading spool file - insurance
2162 against a later release with more variables having written the file.
2164 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2165 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2166 this, as well as the form without the "IPv6" on the front.
2168 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2169 OS/Makefile-Darwin file.
2171 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2172 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2174 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2175 variable. However, some OS use ints for inodes. I've added cast to long int
2176 to get rid of the compiler warning.
2178 19. I had forgotten to lock out "/../" in configuration file names when
2179 ALT_CONFIG_PREFIX was set.
2181 20. Routers used for verification do not need to specify transports. However,
2182 if such a router generated a host list, and callout was configured, Exim
2183 crashed, because it could not find a port number from the (non-existent)
2184 transport. It now assumes port 25 in this circumstance.
2186 21. Added the -t option to exigrep.
2188 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2189 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2190 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2191 them is set, LOOKUP_LSEARCH is forced.
2193 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2195 24. Added sender and host information to the "rejected by local_scan()" log
2196 line; previously there was no indication of these.
2198 25. Added .include_if_exists.
2200 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2201 Linux. It turns out that not all file systems support this. Apparently some
2202 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2203 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2204 not supported on the file descriptor, is now ignored when Exim is trying to
2205 sync a directory. This applies only to Linux.
2207 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2209 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2210 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2213 29. Added support for maildirsize files from supplied patch (modified a bit).
2215 30. The use of :fail: followed by an empty string could lead Exim to respond to
2216 sender verification failures with (e.g.):
2218 550 Verification failed for <xxx>
2219 550 Sender verify failed
2221 where the first response line was missing the '-' that indicates it is not
2222 the final line of the response.
2224 31. The loop for finding the name of the user that called Exim had a hardwired
2225 limit of 10; it now uses the value of finduser_retries, which is used for
2226 all other user lookups.
2228 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2231 33. Exim was neglecting to zero errno before one call of strtol() when
2232 expanding a string and expecting an integer value. On some systems this
2233 resulted in spurious "integer overflow" errors. Also, it was casting the
2234 result into an int without checking.
2236 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2237 did not work. The code looks as if it has *never* worked, though it appears
2238 to have been documented since at least releast 1.62. I have made it work.
2240 35. The "timeout_DNS" error in retry rules, also documented since at least
2241 1.62, also never worked. As it isn't clear exactly what this means, and
2242 clearly it isn't a major issue, I have abolished the feature by treating it
2243 as "timeout", and writing a warning to the main and panic logs.
2245 36. The display of retry rules for -brt wasn't always showing the error code
2248 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2249 timeout_connect_A, timeout_connect_MX.
2251 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2252 to the empty sender.
2254 39. The daemon was not analysing the content of -oX till after it had closed
2255 stderr and disconnected from the controlling terminal. This meant that any
2256 syntax errors were only noted on the panic log, and the return code from
2257 the command was 0. By re-arranging the code a little, I've made the
2258 decoding happen first, so such errors now appear on stderr, and the return
2259 code is 1. However, the actual setting up of the sockets still happens in
2260 the disconnected process, so errors there are still only recorded on the
2263 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2264 connections (as happens on some IP stacks) was logged at start up time as
2265 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2266 it from "IPv6 and IPv4", which means that two separate sockets are being
2269 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2270 says whether AF_INET or AF_INET6 was passed as an argument.
2272 42. Exiwhat output was messed up when time zones were included in log
2275 43. Exiwhat now gives more information about the daemon's listening ports,
2276 and whether -tls-on-connect was used.
2278 44. The "port" option of the smtp transport is now expanded.
2280 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2281 silently ignored. Now an error message is written to the main and panic
2284 46. There's a new ACL modifier called "logwrite" which writes to a log file
2285 as soon as it is encountered.
2287 47. Added $local_user_uid and $local_user_gid at routing time.
2289 48. Exim crashed when trying to verify a sender address that was being
2292 49. Exim was recognizing only a space character after ".include". It now also
2293 recognizes a tab character.
2295 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2296 extracting the relevant information from the specification. The man page no
2297 longer contains scrambled data for the -d option, and I've added a section
2298 at the front about calling Exim under different names.
2300 51. Added "extra_headers" argument to the "mail" command in filter files.
2302 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2305 53. Installed eximstats 1.29.
2307 54. Added transport_filter_timeout as a generic transport option.
2309 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2310 Cc: header lines. This was required by RFC 822, but it not required by RFC
2313 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2314 incoming messages that did not have them. Now it does so only if the
2315 message originates locally, that is, if there is no associated remote host
2316 address. When Resent- header lines are present, this applies to the Resent-
2317 lines rather than the non-Resent- lines.
2319 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2320 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2322 58. Messages for configuration errors now include the name of the main
2323 configuration file - useful now that there may be more than one file in a
2324 list (.included file names were always shown).
2326 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2327 for those rare installations that do not start the daemon as root or run it
2328 setuid root. I've cut out the call to initgroups() if the daemon is not
2331 60. The Exim user and group can now be bound into the binary as text strings
2332 that are looked up at the start of Exim's processing.
2334 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2336 62. Added $mailstore_basename variable.
2338 63. Installed patch to sieve.c from Michael Haardt.
2340 64. When Exim failed to open the panic log after failing to open the main log,
2341 the original message it was trying to log was written to stderr and debug
2342 output, but if they were not available (the usual case in production), it
2343 was lost. Now it is written to syslog before the two lines that record the
2344 failures to open the logs.
2346 65. Users' Exim filters run in subprocesses under the user's uid. It is
2347 possible for a "deliver" command or an alias in a "personal" command to
2348 provoke an address rewrite. If logging of address rewriting is configured,
2349 this fails because the process is not running as root or exim. There may be
2350 a better way of dealing with this, but for the moment (because 4.30 needs
2351 to be released), I have disabled address rewrite logging when running a
2352 filter in a non-root, non-exim process.
2358 1. The buildconfig auxiliary program wasn't quoting the value set for
2359 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2360 not defined. This bug was masked in 4.22 by the effect that was fixed in
2363 2. Some messages that were rejected after a message id was allocated were
2364 shown as "incomplete" by exigrep. It no longer does this for messages that
2365 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2367 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2368 have allow_domain_literals set, the ID did not get logged in the <= line.
2369 Domain literals are now always recognized in Message-ID: header lines.
2371 4. The first argument for a ${extract expansion item is the key name or field
2372 number. Leading and trailing spaces in this item were not being ignored,
2373 causing some misleading effects.
2375 5. When deliver_drop_privilege was set, single queue runner processes started
2376 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2377 same command in the process it spins off) were not dropping privilege.
2379 6. When the daemon running as "exim" started a queue runner, it always
2380 re-executed Exim in the spun-off process. This is a waste of effort when
2381 deliver_drop_privilege is set. The new process now just calls the
2382 queue-runner function directly.
2388 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2391 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2392 ignored. Though the use of -oP was forcing the writing of a pid file, it
2393 was always written to the default place.
2395 3. If the message "no IP address found for host xxxx" is generated during
2396 incoming verification, it is now followed by identification of the incoming
2397 connection (so you can more easily find what provoked it).
2399 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2401 5. Added some features to "harden" Exim a bit more against certain attacks:
2403 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2404 be put in Local/Makefile. This is like the never_users runtime option,
2405 but it cannot be overridden. The default setting is "root".
2407 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2408 prefix string with which any file named in a -C command line option
2411 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2412 is retained for -C and -D only if the caller of Exim is root. Without
2413 it, the exim user may also use -C and -D and retain privilege.
2415 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2416 command line option is disabled.
2418 6. Macro names set by the -D option must start with an upper case letter, just
2419 like macro names defined in the configuration file.
2421 7. Added "dereference=" facility to LDAP.
2423 8. Two instances of the typo "uknown" in the source files are fixed.
2425 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2426 the Configure-Makefile script screwed up while processing it.
2428 10. Incorporated PCRE 4.4.
2430 11. The SMTP synchronization check was not operating right at the start of an
2431 SMTP session. For example, it could not catch a HELO sent before the client
2432 waited for the greeting. There is now a check for outstanding input at the
2433 point when the greeting is written. Because of the duplex, asynchronous
2434 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2435 way, but not yet received, when the check is performed.
2437 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2438 on TCP/IP sockets, because this apparently causes some broken clients to
2441 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2442 unchanged) from the Cygwin maintainer.
2444 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2445 was supported instead of testing for mbx. Effectively a typo.
2447 15. The spa authenticator server code was not checking that the input it
2448 received was valid base64.
2450 16. The debug output line for the "set" modifier in ACLs was not showing the
2451 name of the variable that was being set.
2453 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2455 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2456 lookup on its own. Something else was needed to trigger the lookup. For
2457 example, a match in host_lookup or the need for a host name in a host list.
2458 Now, if $sender_host_name is referenced and the host name has not yet been
2459 looked up, a lookup is performed. If the lookup fails, the variable remains
2460 empty, and $host_lookup_failed is set to "1".
2462 19. Added "eqi" as a case-independent comparison operator.
2464 20. The saslauthd authentication condition could segfault if neither service
2465 nor realm was specified.
2467 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2468 error message that was logged was misleading, and incoming SMTP
2469 connections were dropped. The message is now more accurate, and temporary
2470 errors are given to SMTP connections.
2472 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2473 (except RSET) with a 421 error, until QUIT is received. However, it was
2474 failing to send a response to QUIT.
2476 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2477 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2478 ACL is now run after the helo_try_verify_hosts code.
2480 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2481 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2482 case-independent, so other case variants are also recognized). Apparently
2483 some systems use these upper case variants.
2485 25. If more than two messages were waiting for the same host, and a transport
2486 filter was specified for the transport, Exim sent two messages over the
2487 same TCP/IP connection, and then failed with "socket operation on non-
2488 socket" when it tried to send the third.
2490 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2492 27. The extern definition of crypt16() in expand.c was not being excluded when
2493 the OS had its own crypt16() function.
2495 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2496 as a preferred synonym for return_size_limit, both as an option and as an
2499 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2501 30. Changed the default configuration ACL to relax the local part checking rule
2502 for addresses that are not in any local domains. For these addresses,
2503 slashes and pipe symbols are allowed within local parts, but the sequence
2504 /../ is explicitly forbidden.
2506 31. SPA server authentication was not clearing the challenge buffer before
2509 32. log_message in a "warn" ACL statement was writing to the reject log as
2510 well as to the main log, which contradicts the documentation and doesn't
2511 seem right (because no rejection is happening). So I have stopped it.
2513 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2514 However, I am unable to do any testing of this.
2516 34. Fixed an infelicity in the appendfile transport. When checking directories
2517 for a mailbox, to see if any needed to be created, it was accidentally
2518 using path names with one or more superfluous leading slashes; tracing
2519 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2521 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2522 contents are added to the log line that is written for every discarded
2523 recipient. (Previously a log_message setting was ignored.)
2525 36. The ${quote: operator now quotes the string if it is empty.
2527 37. The install script runs exim in order to find its version number. If for
2528 some reason other than non-existence or emptiness, which it checks, it
2529 could not run './exim', it was installing it with an empty version number,
2530 i.e. as "exim-". This error state is now caught, and the installation is
2533 38. An argument was missing from the function that creates an error message
2534 when Exim fails to connect to the socket for saslauthd authentication.
2535 This could cause Exim to crash, or give a corrupted message.
2537 39. Added isip, isip4, and isip6 to ${if conditions.
2539 40. The ACL variables $acl_xx are now saved with the message, and can be
2540 accessed later in routers, transports, and filters.
2542 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2543 strings in the file are not string-expanded.
2545 42. If a MAIL command specified a SIZE value that was too large to fit into an
2546 int variable, the check against message_size_limit failed. Such values are
2547 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2548 day this will have to be increased, but I don't think I want to be around
2549 when emails are that large.
2556 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2557 iconv() is not standard in FreeBSD.
2559 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2560 IPv6 enabled. The observed symptom was a segmentation fault on return from
2561 the function os_common_find_running_interfaces() in src/os.c.
2563 3. In the check_special_case() function in daemon.c I had used "errno" as an
2564 argument name, which causes warnings on some systems. This was basically a
2565 typo, since it was named "eno" in the comments!
2567 4. The code that waits for the clock to tick (at a resolution of some fraction
2568 of a second) so as to ensure message-id uniqueness was always waiting for
2569 at least one whole tick, when it could have waited for less. [This is
2570 almost certainly not relevant at current processor speeds, where it is
2571 unlikely to ever wait at all. But we try to future-proof.]
2573 5. The function that sleeps for a time interval that includes fractions of a
2574 second contained a race. It did not block SIGALRM between setting the
2575 timer, and suspending (a couple of lines later). If the interval was short
2576 and the sigsuspend() was delayed until after it had expired, the suspension
2577 never ended. On busy systems this could lead to processes getting stuck for
2580 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2581 process, before it forks any delivery processes. The open lookup caching
2582 mechanism meant that the open file or database connection was passed into
2583 the delivery process. The problem was that delivery processes always tidy
2584 up cached lookup data. This could cause a problem for the next delivery
2585 process started by the queue runner, because the external queue runner
2586 process does not know about the closure. So the next delivery process
2587 still has data in the lookup cache. In the case of a file lookup, there was
2588 no problem because closing a file descriptor in a subprocess doesn't affect
2589 the parent. However, if the lookup was caching a connection to a database,
2590 the connection was closed, and the second delivery process was likely to
2591 see errors such as "PGSQL: query failed: server closed the connection
2592 unexpectedly". The problem has been fixed by closing all cached lookups
2593 in a queue runner before running a delivery process.
2595 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2596 seem to have the "const" qualifier which it has on other OS. I've
2599 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2600 *of the same type* (client or server) with the same public name that an
2601 error should be diagnosed.
2603 9. When Exim looked up a host name for an IP address, but failed to find the
2604 original IP address when looking up the host name (a safety check), it
2605 output the message "<ip address> does not match any IP for NULL", which was
2606 confusing, to say the least. The bug was that the host name should have
2607 appeared instead of "NULL".
2609 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2610 user that is built into the binary, and the -C or -D options is used, root
2611 privilege is dropped before the configuration file is read. In addition,
2612 logging is switched to stderr instead of the normal log files. If the
2613 configuration then re-defines the Exim user, the unprivileged environment
2614 is probably not what is expected, so Exim logs a panic warning message (but
2617 However, if deliver_drop_privilege is set, the unprivileged state may well
2618 be exactly what is intended, so the warning has been cut out in that case,
2619 and Exim is allowed to try to write to its normal log files.
2625 1. smtp_return_error_details was not giving details for temporary sender
2626 or receiver verification errors.
2628 2. Diagnose a configuration error if two authenticators have the same public
2631 3. Exim used not to create the message log file for a message until the first
2632 delivery attempt. This could be confusing when incoming messages were held
2633 for policy or load reasons. The message log file is now created at the time
2634 the message is received, and an initial "Received" line is written to it.
2636 4. The automatically generated man page for command line options had a minor
2637 bug that caused no ill effects; however, a more serious problem was that
2638 the procedure for building the man page automatically didn't always
2639 operate. Consequently, release 4.20 contains an out-of-date version. This
2640 shouldn't happen again.
2642 5. When building Exim with embedded Perl support, the script that builds the
2643 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2644 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2646 6. The freeze_tell option was not being used for messages that were frozen on
2647 arrival, either by an ACL or by local_scan().
2649 7. Added the smtp_incomplete_transaction log selector.
2651 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2652 was accepting AUTH without a new EHLO.
2654 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2655 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2658 10. Logging of TCP/IP connections (when configured) now happens in the main
2659 daemon process instead of the child process, so that the TCP/IP connection
2660 count is more accurate (but it can never be perfect).
2662 11. The use of "drop" in a nested ACL was not being handled correctly in the
2663 outer ACL. Now, if condition failure induced by the nested "drop" causes
2664 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2665 or "require"), the connection is dropped.
2667 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2668 that yield "discard" can now be used with an "accept" or a "discard" verb,
2669 but an error is generated for any others (because I can't see a useful way
2670 to define what should happen).
2672 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2673 are now processed in the same way as lines in the Exim configuration file.
2674 In particular, continuation lines are supported.
2676 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2678 15. Added -ti meaning -t -i.
2680 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2681 domains, and warn by logging if others are found.
2683 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2684 returned from GIFCONF when Exim is trying to find the list of interfaces on
2685 a host. The code in os.c has been modified to copy each ifreq to an aligned
2686 structure in all cases.
2688 Also, in some cases, the returned ifreq's were being copied to a 'struct
2689 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2690 means the last couple of bytes of an IPv6 address could be chopped if the
2691 ifreq contained only a normal sockaddr (14 bytes storage).
2693 18. Named domain lists were not supported in the hosts_treat_as_local option.
2694 An entry such as +xxxx was not recognized, and was treated as a literal
2697 19. Ensure that header lines added by a DATA ACL are included in the reject log
2698 if the ACL subsequently rejects the message.
2700 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2701 MD5 (which is deprecated).
2703 21. When testing a filter file using -bf, Exim was writing a message when it
2704 took the sender from a "From " line in the message, but it was not doing so
2705 when it took $return_path from a Return-Path: header line. It now does.
2707 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2708 with a valid header line field name (a series of printing characters
2709 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2711 23. Changed "disc" in the source to "disk" to conform to the documentation and
2712 the book and for uniformity.
2714 24. Ignore Sendmail's -Ooption=value command line item.
2716 25. When execve() failed while trying to run a command in a pipe transport,
2717 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2718 could be confused with a return value of 69 from the command itself. This
2719 has been changed to 127, the value the shell returns if it is asked to run
2720 a non-existent command. The wording for the related log line suggests a
2721 non-existent command as the problem.
2723 26. If received_header_text expands to an empty string, do not add a Received:
2724 header line to the message. (Well, it adds a token one on the spool, but
2725 marks it "old" so that it doesn't get used or transmitted.)
2727 27. Installed eximstats 1.28 (addition of -nt option).
2729 28. There was no check for failure on the call to getsockname() in the daemon
2730 code. This can fail if there is a shortage of resources on the system, with
2731 ENOMEM, for example. A temporary error is now given on failure.
2733 29. Contrary to the C standard, it seems that in some environments, the
2734 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2735 program. Exim now does this explicitly; it affects the formatting of
2736 timestamps using strftime().
2738 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2739 complaints. I've now initialized all the variables, to avoid this.
2741 32. Header lines added by a system filter were not being "seen" during
2742 transport-time rewrites.
2744 33. The info_callback() function passed to OpenSSL is set up with type void
2745 (*)(SSL *, int, int), as described somewhere. However, when calling the
2746 function (actually a macro) that sets it up, the type void(*)() is
2747 expected. I've put in a cast to prevent warnings from picky compilers.
2749 34. If a DNS black list lookup found a CNAME record, but there were no A
2750 records associated with the domain it pointed at, Exim crashed.
2752 35. If a DNS black list lookup returned more than one A record, Exim ignored
2753 all but the first. It now scans all returned addresses if a particular IP
2754 value is being sought. In this situation, the contents of the
2755 $dnslist_value variable are a list of all the addresses, separated by a
2758 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2759 to accept a host name and all its aliases if the forward lookup for any of
2760 them yielded the IP address of the incoming connection. Now it accepts only
2761 those names whose forward lookup yields the correct IP address. Any other
2762 names are discarded. This closes a loophole whereby a rogue DNS
2763 administrator could create reverse DNS records to break through a
2764 wildcarded host restriction in an ACL.
2766 37. If a user filter or a system filter that ran in a subprocess used any of
2767 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2768 the wrong values were passed to the pipe command ($thisaddress had the
2769 value of $0, $0 had the value of $1, etc). This bug was introduced by
2770 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2772 38. Improved the line breaking for long SMTP error messages from ACLs.
2773 Previously, if there was no break point between 40 and 75 characters, Exim
2774 left the rest of the message alone. Two changes have been made: (a) I've
2775 reduced the minimum length to 35 characters; (b) if it can't find a break
2776 point between 35 and 75 characters, it looks ahead and uses the first one
2777 that it finds. This may give the occasional overlong line, but at least the
2778 remaining text gets split now.
2780 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2781 file descriptors might be low, and that setting 1000 would always raise it.
2782 It turns out that in some environments, the limit is already over 1000 and
2783 that lowering it causes trouble. So now Exim takes care not to decrease it.
2785 40. When delivering a message, the value of $return_path is set to $sender_
2786 address at the start of routing (routers may change the value). By an
2787 oversight, this default was not being set up when an address was tested by
2788 -bt or -bv, which affected the outcome if any router or filter referred to
2791 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2792 log only when "message" or "log_message" are set. However, if one of the
2793 conditions was an address verification, or a call to a nested ACL, the
2794 messages generated by the underlying test were being passed through. This
2795 no longer happens. The underlying message is available in $acl_verify_
2796 message for both "message" and "log_message" expansions, so it can be
2797 passed through if needed.
2799 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2800 new expansion $bh_ to give the encoded byte string without charset
2801 translation. Translation happens only if iconv() is available; HAVE_ICONV
2802 indicates this at build time. HEADERS_CHARSET gives the charset to
2803 translate to; headers_charset can change it in the configuration, and
2804 "headers charset" can change it in an individual filter file.
2806 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
2807 that creates RFC 2047 encoded "words" labels them as that charset instead
2808 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2809 expansion operator; (ii) when Exim creates a From: line for a local
2810 message; (iii) when a header line is rewritten to include a "phrase" part.
2812 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2813 buggy, causing it to skip the first lines of messages whose message ID
2814 ended in 'D'. This would not have bitten before Exim release 4.14, because
2815 message IDs were unlikely to end in 'D' before then. The effect was to have
2816 incorrect size information for certain domains.
2818 45. #include "config.h" was missing at the start of the crypt16.c module. This
2819 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2822 46. If there was a timeout during a "random" callout check, Exim treated it as
2823 a failure of the random address, and carried on sending RSET and the real
2824 address. If the delay was just some slowness somewhere, the response to the
2825 original RCPT would be taken as a response to RSET and so on, causing
2826 mayhem of various kinds.
2828 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2829 when I implemented it. It didn't allow for the fact that some option values
2830 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2831 the right test for positive values.
2833 48. Domain names in DNS records are case-independent. Exim always looks them up
2834 in lower case. Some resolvers return domain names in exactly the case they
2835 appear in the zone file, that is, they may contain uppercase letters. Not
2836 all resolvers do this - some return always lower case. Exim was treating a
2837 change of case by a resolver as a change of domain, similar to a widening
2838 of a domain abbreviation. This triggered its re-routing code and so it was
2839 trying to route what was effectively the same domain again. This normally
2840 caused routing to fail (because the router wouldn't handle the domain
2841 twice). Now Exim checks for this case specially, and just changes the
2842 casing of the domain that it ultimately uses when it transmits the message
2845 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
2848 50. If a filter generated a file delivery with a non-absolute name (possible if
2849 no home directory exists for the router), the forbid_file option was not
2852 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
2853 the existing equality matching.
2855 52. Exim was using ints instead of ino_t variables in some places where it was
2856 dealing with inode numbers.
2858 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
2859 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
2860 called TMPDIR, and if it finds it is different, it changes its value.
2862 54. The smtp_printf() function is now made available to local_scan() so
2863 additional output lines can be written before returning. There is also an
2864 smtp_fflush() function to enable the detection of a dropped connection.
2865 The variables smtp_input and smtp_batched_input are exported to
2868 55. Changed the default runtime configuration: the message "Unknown user"
2869 has been removed from the ACL, and instead placed on the localuser router,
2870 using the cannot_route_message feature. This means that any verification
2871 failures that generate their own messages won't get overridden. Similarly,
2872 the "Unrouteable address" message that was in the ACL for unverifiable
2873 relay addresses has also been removed.
2875 56. Added hosts_avoid_esmtp to the smtp transport.
2877 57. The exicyclog script was not checking for the esoteric option
2878 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
2879 will work only if exicyclog is run under the appropriate euid.
2881 58. Following a discussion on the list, the rules by which Exim recognises line
2882 endings on incoming messages have been changed. The -dropcr and drop_cr
2883 options are now no-ops, retained only for backwards compatibility. The
2884 following line terminators are recognized: LF CRLF CR. However, special
2885 processing applies to CR:
2887 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
2888 nor a local message in the state where . is a terminator.
2890 (ii) If a bare CR is encountered in a header line, an extra space is added
2891 after the line terminator so as not to end the header. The reasoning
2892 behind this is that bare CRs in header lines are most likely either
2893 to be mistakes, or people trying to play silly games.
2895 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
2896 was being incorrectly given as 18 bytes larger than it should have been.
2897 This is a VOB (very old bug).
2899 60. This may never have affected anything current, but just in case it has:
2900 When the local host is found other than at the start of a list of hosts,
2901 the local host, those with the same MX, and any that follow, are discarded.
2902 When the list in question was part of a longer list of hosts, the following
2903 hosts (not currently being processed) were also being discarded. This no
2904 longer happens. I'm not sure if this situation could ever has previously
2907 61. Added the "/MX" feature to lists of hosts in the manualroute and query
2910 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
2911 header. This is something that is recommended in a new Internet Draft, and
2912 is something that is documented as being done by Sendmail. There are two
2913 possible values. For messages generated by the autoreply transport, Exim
2916 Auto-Submitted: auto-replied
2918 whereas for all other generated messages (e.g. bounces) it adds
2920 Auto-Submitted: auto-generated
2922 63. The "personal" condition in filters now includes a test for the
2923 Auto-Submitted: header. If it contains the string "auto-" the message it
2924 not considered personal.
2926 64. Added rcpt_include_affixes as a generic transport option.
2928 65. Added queue_only_override (default true).
2930 66. Added the syslog_duplication option.
2932 67. If what should have been the first header line of a message consisted of
2933 a space followed by a colon, Exim was mis-interpreting it as a header line.
2934 It isn't of course - it is syntactically invalid and should therefore be
2935 treated as the start of the message body. The misbehaviour could have
2936 caused a number of strange effects, including loss of data in subsequent
2937 header lines, and spool format errors.
2939 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
2940 client host had authenticated. This control can now be exercised by an ACL
2941 for more flexibility.
2943 69. By default, callouts do not happen when testing with -bh. There is now a
2944 variant, -bhc, which does actually run the callout code, including
2945 consulting and updating the callout cache.
2947 70. Added support for saslauthd authentication, courtesy of Alexander
2950 71. If statvfs() failed on the spool or log directories while checking their
2951 size for availability, Exim confusingly gave the error "space shortage".
2952 Furthermore, in debugging mode it crashed with a floating point exception.
2953 These checks are done if check_{spool,log}_{space,inodes} are set, and when
2954 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
2955 serious problem, Exim now writes to the main and panic logs when this
2956 happens, with details of the failure. It then refuses to accept the
2957 incoming message, giving the message "spool directory problem" or "log
2958 directory problem" with a 421 code for SMTP messages.
2960 72. When Exim is about to re-exec itself, it ensures that the file descriptors
2961 0, 1, and 2 exist, because some OS complain for execs without them (see
2962 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
2963 descriptors. However, the code omitted to check that the open succeeded,
2964 causing mysterious errors if for some reason the permissions on /dev/null
2965 got screwed. Now Exim writes a message to the main and panic logs, and
2966 bombs out if it can't open /dev/null.
2968 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
2969 interact so that it is all more flexible. It is supposed to remain
2970 backwards compatible. Also added extra_local_interfaces.
2972 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
2973 to bomb out with an assertion failure - to the client this appears as a
2974 connection drop. This problem occurs in the part of the code that was taken
2975 from the Samba project. Fortunately, the assertion is in a very simple
2976 function, so I have fixed this by reproducing the function inline in the
2977 one place where it is called, and arranging for authentication to fail
2978 instead of killing the process with assert().
2980 75. The SPA client code was not working when the server requested OEM rather
2981 than Unicode encoding.
2983 76. Added code to make require_files with a specific uid setting more usable in
2984 the case where statting the file as root fails - usually a non-root-mounted
2985 NFS file system. When this happens and the failure is EACCES, Exim now
2986 forks a subprocess and does the per-uid checking as the relevant uid.
2988 77. Added process_log_path.
2990 78. If log_file_path was not explicitly set, a setting of check_log_space or
2991 check_log_inodes was ignored.
2993 79. If a space check for the spool or log partitions fails, the incident is now
2994 logged. Of course, in the latter case the data may get lost...
2996 80. Added the %p formatting code to string_format() so that it can be used to
2997 print addresses in debug_print(). Adjusted all the address printing in the
2998 debugging in store.c to use %p rather than %d.
3000 81. There was a concern that a line of code in smtp_in.c could overflow a
3001 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3002 initially expressed, the concern was not well-founded, because trailing
3003 spaces are removed early. However, if the trailing spaces were followed by
3004 a NULL, they did not get removed, so the overflow was possible. Two fixes
3007 (a) I re-wrote the offending code in a cleaner fashion.
3008 (b) If an incoming SMTP command contains a NULL character, it is rejected
3011 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3012 runs initgroups(), so that if the Exim user is in any additional groups,
3013 they will be used during message reception.
3019 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob