Put in some casts for signed/unsigned character strings that Tom forgot
[exim.git] / doc / doc-txt / ChangeLog
294520c8 1$Cambridge: exim/doc/doc-txt/ChangeLog,v 1.159 2005/06/16 20:02:42 tom Exp $
3Change log file for Exim from version 4.21
bbe902f0 6
7Exim version 4.52
10TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
12PH/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.
17TK/01 Fix poll() being unavailable on Mac OSX 10.2.
19PH/02 Reduce the amount of output that "make" produces by default. Full output
20 can still be requested.
22PH/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.
27TF/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.
31PH/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
34 in the header line.
36TF/03 Added the control = fakedefer ACL modifier.
38TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
39 Mark Lowes for thorough testing.
870f6ba8 40
41TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
43TK/03 Merged latest SRS patch from Miles Wilton.
45PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
46 with the definition in sysexits.h (which is #included earlier).
47 Fortunately, Exim does not actually use EX_OK. The code used to try to
48 preserve the sysexits.h value, by assumimg that macro definitions were
49 scanned for macro replacements. I have been disabused of this notion,
50 so now the code just undefines EX_OK before #including unistd.h.
11d337a4 51
52PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
53 in the smtp transport. When a block could not be written in a single
54 write() function, the timeout was being re-applied to each part-write.
55 This seems wrong - if the receiver was accepting one byte at a time it
56 would take for ever. The timeout is now adjusted when this happens. It
57 doesn't have to be particularly precise.
59TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
60 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
62PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
63 without a domain if the check to <postmaster@domain> fails.
65SC/01 Eximstats: added -xls and the ability to specify output files
66 (patch written by Frank Heydlauf).
68SC/02 Eximstats: use FileHandles for outputing results.
70SC/03 Eximstats: allow any combination of xls, txt, and html output.
72SC/04 Eximstats: fixed display of large numbers with -nvr option
74SC/05 Eximstats: fixed merging of reports with empty tables.
76SC/06 Eximstats: added the -include_original_destination flag
78SC/07 Eximstats: removed tabs and trailing whitespace.
80TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
82TK/06 MBOX spool code: Add real "From " MBOX separator line
83 so the .eml file is really in mbox format (even though
84 most programs do not really care). Patch from Alex Miller.
86TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
87 The latter is generated from $received_to and is only set if the
88 message has one envelope recipient. SA can use these headers,
89 obviously out-of-the-box. Patch from Alex Miller.
91PH/08 The ${def test on a variable was returning false if the variable's
92 value was "0", contrary to what the specification has always said!
93 The result should be true unless the variable is empty.
95PH/09 The syntax error of a character other than { following "${if
96 def:variable_name" (after optional whitespace) was not being diagnosed.
97 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
98 accidental colon was present, for example, could give incorrect results.
100PH/10 Tidied the code in a number of places where the st_size field of a stat()
101 result is used (not including appendfile, where other changes are about
102 to be made).
104PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
105 This involved changing a lot of size variables from int to off_t. It
106 should work with maildirs and everything.
108TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
109 spamd dying while we are connected to it.
111TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
112 <jgh@wizmail.org>
114PH/12 Applied Alex Kiernan's patch for the API change for the error callback
115 function for BDB 4.3.
117PH/13 Changed auto_thaw such that it does not apply to bounce messages.
119PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
120 the sources for PCRE have been re-arranged and more files are now
121 involved.
123PH/15 The code I had for printing potentially long long variables in PH/11
124 above was not the best (it lost precision). The length of off_t variables
125 is now inspected at build time, and an appropriate printing format (%ld
126 or %lld) is chosen and #defined by OFF_T_FMT. We also define ASSUME_
127 LONG_LONG_SUPPORT if the length is greater than 4. This is needed for the
128 internal formatting function string_vformat().
130PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
131 the configuration file to be ":syslog", then the script "guesses" where
132 the logs files are, rather than using the compiled in default. In our
133 case the guess is not the same as the compiled default, so the script
134 suddenly stopped working when I started to use syslog. The patch checks
135 to see if log_file_path is "". If so, it attempts to read it from exim
136 with no configuration file to get the compiled in version, before it
137 falls back to the previous guessing code."
139TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
140 implementing BATV in an Exim configuration. See NewStuff for the gory
141 details.
e5a9dba6 143
144Exim version 4.51
147TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
148 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2f079f46 150TK/02 Fix ACL "control" statement not being available in MIME ACL.
152TK/03 Fix ACL "regex" condition not being available in MIME ACL.
154PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
155 to test Sieve filters that use "vacation".
157PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
158 that changes the way the GnuTLS parameters are stored in the cache file.
159 The new format can be generated externally. For backward compatibility,
160 if the data in the cache doesn't make sense, Exim assumes it has read an
161 old-format file, and it generates new data and writes a new file. This
162 means that you can't go back to an older release without removing the
163 file.
165PH/03 A redirect router that has both "unseen" and "one_time" set does not
166 work if there are any delivery delays because "one_time" forces the
167 parent to be marked "delivered", so its unseen clone is never tried
168 again. For this reason, Exim now forbids the simultaneous setting of
169 these two options.
171PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
172 redirected to themselves ("homonym" addresses). Read the long ChangeLog
173 entry if you want to know the details. The fix, however, neglected to
174 consider the case when local delivery batching is involved. The test for
175 "previously delivered" was not happening when checking to see if an
176 address could be batched with a previous (undelivered) one; under
177 certain circumstances this could lead to multiple deliveries to the same
c2c19e9d 178 address.
180PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
181 in its include files, and this causes problems building Exim.
183PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
184 header_syntax) but Exim was just ignoring anything given after a slash.
185 In particular, this caused confusion with an attempt to use "verify =
186 reverse_host_lookup/defer_ok". An error is now given when options are
187 supplied for verify items that do not have them. (Maybe reverse_host_
188 lookup should have a defer_ok option, but that's a different point.)
190PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
191 defined by RFC 821) to 2048, because there were problems with some AUTH
192 commands, and RFC 1869 says the size should be increased for extended
193 SMTP commands that take arguments.
195PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
196 Finch).
198PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
199 "unknown" error; now it says that the functionality isn't in the binary.
8d67ada3 200
201PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
202 an address' error message when a string expansion fails (syntax or
203 whatever). Otherwise the password may appear in the log. Following change
204 PH/42 below, there is no longer a chance of it appearing in a bounce
205 message.
49c2d5ea 206
207PH/11 Installed exipick version 20050225.0 from John Jetmore.
209PH/12 If the last host in a fallback_hosts list was multihomed, only the first
210 of its addresses was ever tried. (Bugzilla bug #2.)
212PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
213 the result incorrectly in the debug output. (It correctly added a newline
214 to what was transported.)
216TF/01 Added $received_time.
218PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
219 commented out examples of how to interface to a virus scanner and to
220 SpamAssassin. Also added commented examples of av_scanner and
221 spamd_address settings.
223PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
224 and controls are allowed in which ACLs. There were a couple of minor
225 errors. Some of the entries in the conditions table (which is a table of
226 where they are NOT allowed) were getting very unwieldy; rewrote them as a
227 negation of where the condition IS allowed.
229PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
231PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
232 header file does not have a version number, so I've had to invent a new
233 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
234 API. The code is untested by me (my Linux distribution still has 0.3.2 of
235 radiusclient), but it was contributed by a Radius user.
237PH/18 Installed Lars Mainka's patch for the support of CRL collections in
238 files or directories, for OpenSSL.
240PH/19 When an Exim process that is running as root has to create an Exim log
241 file, it does so in a subprocess that runs as exim:exim so as to get the
242 ownership right at creation (otherwise, other Exim processes might see
243 the file with the wrong ownership). There was no test for failure of this
244 fork() call, which would lead to the process getting stuck as it waited
245 for a non-existent subprocess. Forks do occasionally fail when resources
246 run out. I reviewed all the other calls to fork(); they all seem to check
247 for failure.
249PH/20 When checking for unexpected SMTP input at connect time (before writing
250 the banner), Exim was not dealing correctly with a non-positive return
251 from the read() function. If the client had disconnected by this time,
252 the result was a log entry for a synchronization error with an empty
253 string after "input=" when read() returned zero. If read() returned -1
254 (an event I could not check), uninitialized data bytes were printed.
255 There were reports of junk text (parts of files, etc) appearing after
256 "input=".
258PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
260PH/22 Added support for macro redefinition, and (re)definition in between
261 driver and ACL definitions.
263PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
264 forgetting to use the resulting value; it was using the unexpanded value.
266PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
267 hadn't been configured. The fix is from Juergen Kreileder, who
268 understands it better than I do:
270 "Here's what I see happening with three configured cyrus_sasl
271 authenticators configured (plain, login, cram-md5):
273 On startup auth_cyrus_sasl_init() gets called for each of these.
274 This means three calls to sasl_listmech() without a specified mech_list.
275 => SASL tests which mechs of all available mechs actually work
276 => three warnings about OTP not working
277 => the returned list contains: plain, login, cram-md5, digest-md5, ...
279 With the patch, sasl_listmech() also gets called three times. But now
280 SASL's mech_list option is set to the server_mech specified in the the
281 authenticator. Or in other words, the answer from sasl_listmech()
282 gets limited to just the mech you're testing for (which is different
283 for each call.)
284 => the return list contains just 'plain' or 'login', 'cram-md5' or
285 nothing depending on the value of ob->server_mech.
287 I've just tested the patch: Authentication still works fine,
288 unavailable mechs specified in the exim configuration are still
289 caught, and the auth.log warnings about OTP are gone."
291PH/25 When debugging is enabled, the contents of the command line are added
292 to the debugging output, even when log_selector=+arguments is not
293 specified.
295PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
296 answer is "GNU", and only if the return is "GNU/something" is the answer
297 "Linux".
299PH/27 $acl_verify_message is now set immediately after the failure of a
300 verification in an ACL, and so is available in subsequent modifiers. In
301 particular, the message can be preserved by coding like this:
303 warn !verify = sender
304 set acl_m0 = $acl_verify_message
306 Previously, $acl_verify_message was set only while expanding "message"
307 and "log_message" when a very denied access.
309PH/28 Modified OS/os.c-Linux with
311 -#ifndef OS_LOAD_AVERAGE
312 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
314 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
315 nomenclature these days.)
317PH/29 Installed patch from the Sieve maintainer that adds the options
318 sieve_useraddress and sieve_subaddress to the redirect router.
320PH/30 In these circumstances:
321 . Two addresses routed to the same list of hosts;
322 . First host does not offer TLS;
323 . First host accepts first address;
324 . First host gives temporary error to second address;
325 . Second host offers TLS and a TLS session is established;
326 . Second host accepts second address.
327 Exim incorrectly logged both deliveries with the TLS parameters (cipher
328 and peerdn, if requested) that were in fact used only for the second
329 address.
7e8bec7a 330
331PH/31 When doing a callout as part of verifying an address, Exim was not paying
332 attention to any local part prefix or suffix that was matched by the
333 router that accepted the address. It now behaves in the same way as it
334 does for delivery: the affixes are removed from the local part unless
335 rcpt_include_affixes is set on the transport.
337PH/32 Add the sender address, as F=<...>, to the log line when logging a
338 timeout during the DATA phase of an incoming message.
340PH/33 Sieve envelope tests were broken for match types other than :is. I have
341 applied a patch sanctioned by the Sieve maintainer.
c688b954 342
343PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
344 the uid or gid is negative. A case of a negative gid caused this to be
345 noticed. The fix allows for either to be negative.
347PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
348 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
349 overlooked.
351PH/36 The change PH/12 above was broken. Fixed it.
353PH/37 Exim used to check for duplicate addresses in the middle of routing, on
354 the grounds that routing the same address twice would always produce the
355 same answer. This might have been true once, but it is certainly no
356 longer true now. Routing a child address may depend on the previous
357 routing that produced that child. Some complicated redirection strategies
358 went wrong when messages had multiple recipients, and made Exim's
359 behaviour dependent on the order in which the addresses were given.
361 I have moved the duplicate checking until after the routing is complete.
362 Exim scans the addresses that are assigned to local and remote
363 transports, and removes any duplicates. This means that more work will be
364 done, as duplicates will always all be routed, but duplicates are
365 presumably rare, so I don't expect this is of any significance.
367 For deliveries to pipes, files, and autoreplies, the duplicate checking
368 still happens during the routing process, since they are not going to be
369 routed further.
371PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
372 It corrects a timeout issue with spamd. This is Ian's comment: "The
373 background is that sometimes spamd either never reads data from a
374 connection it has accepted, or it never writes response data. The exiscan
375 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
376 blindly assumes that writes won't block so it may never time out."
378PH/39 Allow G after quota size as well as K and M.
380PH/40 The value set for $authenticated_id in an authenticator may not contain
381 binary zeroes or newlines because the value is written to log lines and
382 to spool files. There was no check on this. Now the value is run through
383 the string_printing() function so that such characters are converted to
384 printable escape sequences.
386PH/41 $message_linecount is a new variable that contains the total number of
387 lines in the message. Compare $body_linecount, which is the count for the
388 body only.
390PH/42 Exim no longer gives details of delivery errors for specific addresses in
391 bounce and delay warning messages, except in certain special cases, which
392 are as follows:
394 (a) An SMTP error message from a remote host;
395 (b) A message specified in a :fail: redirection;
396 (c) A message specified in a "fail" command in a system filter;
397 (d) A message specified in a FAIL return from the queryprogram router;
398 (e) A message specified by the cannot_route_message router option.
400 In these cases only, Exim does include the error details in bounce and
401 warning messages. There are also a few cases where bland messages such
402 as "unrouteable address" or "local delivery error" are given.
404PH/43 $value is now also set for the "else" part of a ${run expansion.
406PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
407 being worked on, but at least Exim now implements the latest version to
408 play with."
410PH/45 In a pipe transport, although a timeout while waiting for the pipe
411 process to complete was treated as a delivery failure, a timeout while
412 writing the message to the pipe was logged, but erroneously treated as a
413 successful delivery. Such timeouts include transport filter timeouts. For
414 consistency with the overall process timeout, these timeouts are now
415 treated as errors, giving rise to delivery failures by default. However,
416 there is now a new Boolean option for the pipe transport called
417 timeout_defer, which, if set TRUE, converts the failures into defers for
418 both kinds of timeout. A transport filter timeout is now identified in
419 the log output.
421PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
422 systems where "make" and "gmake" are different, calling "gmake" at top
423 level broke things. I've arranged for the value of $(MAKE) to be passed
424 from the Makefile to this script so that it can call the same version of
425 "make".
7982096b 427
428A note about Exim versions 4.44 and 4.50
431Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
432changes of various kinds. As a consequence, a big documentation update was
433needed. This delayed the release for rather longer than seemed good, especially
434in the light of a couple of (minor) security issues. Therefore, the changes
435that fixed bugs were backported into 4.43, to create a 4.44 maintenance
436release. So 4.44 and 4.50 are in effect two different branches that both start
437from 4.43.
439I have left the 4.50 change log unchanged; it contains all the changes since
4404.43. The change log for 4.44 is below; many of its items are identical to
441those for 4.50. This seems to be the most sensible way to preserve the
442historical information.
f7b63901 445Exim version 4.50
448 1. Minor wording change to the doc/README.SIEVE file.
139059f6 450 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5fe762f6 451 computation of the current number of files was incorrect.
495ae4b0 452
453 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
454 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
455 place.
457 4. Give more explanation in the error message when the command for a transport
458 filter fails to execute.
460 5. There are several places where Exim runs a non-Exim command in a
461 subprocess. The SIGUSR1 signal should be disabled for these processes. This
462 was being done only for the command run by the queryprogram router. It is
463 now done for all such subprocesses. The other cases are: ${run, transport
464 filters, and the commands run by the lmtp and pipe transports.
466 6. Added CONFIGURE_GROUP build-time option.
468 7. Some older OS have a limit of 256 on the maximum number of file
469 descriptors. Exim was using setrlimit() to set 1000 as a large value
470 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
471 systems. I've change it so that if it can't get 1000, it tries for 256.
35edf2ff 472
473 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
474 was an oversight, and furthermore, ever since the addition of extra
475 controls (e.g. 4.43/32), the checks on when to allow different forms of
476 "control" were broken. There should now be diagnostics for all cases when a
477 control that does not make sense is encountered.
479 9. Added the /retain_sender option to "control=submission".
48110. $recipients is now available in the predata ACL (oversight).
48311. Tidy the search cache before the fork to do a delivery from a message
484 received from the command line. Otherwise the child will trigger a lookup
485 failure and thereby defer the delivery if it tries to use (for example) a
486 cached ldap connection that the parent has called unbind on.
48812. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
489 of $address_data from the recipient verification was clobbered by the
490 sender verification.
49213. The value of address_data from a sender verification is now available in
493 $sender_address_data in subsequent conditions in the ACL statement.
49514. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
49715. Added a new option "connect=<time>" to callout options, to set a different
498 connection timeout.
50016. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
501 was its contents. (It was OK if the option was not defined at all.)
50317. A "Completed" log line is now written for messages that are removed from
504 the spool by the -Mrm option.
50618. New variables $sender_verify_failure and $recipient_verify_failure contain
507 information about exactly what failed.
50919. Added -dd to debug only the daemon process.
51120. Incorporated Michael Haardt's patch to ldap.c for improving the way it
512 handles timeouts, both on the server side and network timeouts. Renamed the
513 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
51521. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
516 It is now set to "smtps".
51822. $host_address is now set to the target address during the checking of
519 ignore_target_hosts.
52123. When checking ignore_target_hosts for an ipliteral router, no host name was
522 being passed; this would have caused $sender_host_name to have been used if
523 matching the list had actually called for a host name (not very likely,
524 since this list is usually IP addresses). A host name is now passed as
525 "[x.x.x.x]".
52724. Changed the calls that set up the SIGCHLD handler in the daemon to use the
528 code that specifies a non-restarting handler (typically sigaction() in
529 modern systems) in an attempt to fix a rare and obscure crash bug.
53125. Narrowed the window for a race in the daemon that could cause it to ignore
532 SIGCHLD signals. This is not a major problem, because they are used only to
533 wake it up if nothing else does.
53526. A malformed maildirsize file could cause Exim to calculate negative values
536 for the mailbox size or file count. Odd effects could occur as a result.
537 The maildirsize information is now recalculated if the size or filecount
538 end up negative.
54027. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
541 support for a long time. Removed HAVE_SYS_VFS_H.
54328. Installed the latest version of exipick from John Jetmore.
54529. In an address list, if the pattern was not a regular expression, an empty
546 subject address (from a bounce message) matched only if the pattern was an
547 empty string. Non-empty patterns were not even tested. This was the wrong
548 because it is perfectly reasonable to use an empty address as part of a
549 database query. An empty address is now tested by patterns that are
550 lookups. However, all the other forms of pattern expect the subject to
551 contain a local part and a domain, and therefore, for them, an empty
552 address still always fails if the pattern is not itself empty.
55430. Exim went into a mad DNS loop when attempting to do a callout where the
555 host was specified on an smtp transport, and looking it up yielded more
556 than one IP address.
55831. Re-factored the code for checking spool and log partition space into a
559 function that finds that data and another that does the check. The former
560 is then used to implement four new variables: $spool_space, $log_space,
561 $spool_inodes, and $log_inodes.
56332. The RFC2047 encoding function was originally intended for short strings
564 such as real names; it was not keeping to the 75-character limit for
565 encoded words that the RFC imposes. It now respects the limit, and
566 generates multiple encoded words if necessary. To be on the safe side, I
567 have increased the buffer size for the ${rfc2047: expansion operator from
568 1024 to 2048 bytes.
57033. It is now permitted to omit both strings after an "if" condition; if the
571 condition is true, the result is "true". As before, when the second string
572 is omitted, a false condition yields an empty string. This makes it less
573 cumbersome to write custom ACL and router conditions.
57534. Failure to deliver a bounce message always caused it to be frozen, even if
576 there was an errors_to setting on the router. The errors_to setting is now
577 respected.
57935. If an IPv6 address is given for -bh or -bhc, it is now converted to the
580 canonical form (fully expanded) before being placed in
581 $sender_host_address.
58336. The table in the code that translates DNS record types into text (T_A to
584 "A" for instance) was missing entries for NS and CNAME. It is just possible
585 that this could have caused confusion if both these types were looked up
586 for the same domain, because the text type is used as part of Exim's
587 per-process caching. But the chance of anyone hitting this buglet seems
588 very small.
59037. The dnsdb lookup has been extended in a number of ways.
592 (1) There is a new type, "zns", which walks up the domain tree until it
593 finds some nameserver records. It should be used with care.
595 (2) There is a new type, "mxh", which is like "mx" except that it returns
596 just the host names, not the priorities.
598 (3) It is now possible to give a list of domains (or IP addresses) to be
599 looked up. The behaviour when one of the lookups defers can be
600 controlled by a keyword.
7bb56e1f 601
ea3bc19b 602 (4) It is now possible to specify the separator character for use when
7bb56e1f 603 multiple records are returned.
33397d19 604
60538. The dnslists ACL condition has been extended: it is now possible to supply
606 a list of IP addresses and/or domains to be looked up in a particular DNS
607 domain.
60939. Added log_selector=+queue_time_overall.
61140. When running the queue in the test harness, wait just a tad after forking a
612 delivery process, to get repeatability of debugging output.
61441. Include certificate and key file names in error message when GnuTLS fails
615 to set them up, because the GnuTLS error message doesn't include the name
616 of the failing file when there is a problem reading it.
61842. Allow both -bf and -bF in the same test run.
62043. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
62244. The "Exiscan patch" is now merged into the mainline Exim source.
62445. Sometimes the final signoff response after QUIT could fail to get
625 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
626 < 0 before doing a fflush(). This bug looks as though it goes back to the
627 introduction of TLS in release 3.20, but "sometimes" must have been rare
628 because the tests only now provoked it.
63046. Reset the locale to "C" after calling embedded Perl, in case it was changed
631 (this can affect the format of dates).
63347. exim_tidydb, when checking for the continued existence of a message for
634 which it has found a message-specific retry record, was not finding
635 messages that were in split spool directories. Consequently, it was
636 deleting retry records that should have stayed in existence.
63848. Steve fixed some bugs in eximstats.
64049. The SPA authentication driver was not abandoning authentication and moving
641 on to the next authenticator when an expansion was forced to fail,
642 contradicting the general specification for all authenticators. Instead it
643 was generating a temporary error. It now behaves as specified.
64550. The default ordering of permitted cipher suites for GnuTLS was pessimal
646 (the order specifies the preference for clients). The order is now AES256,
647 AES128, 3DES, ARCFOUR128.
64951. Small patch to Sieve code - explicitly set From: when generating an
650 autoreply.
65252. Exim crashed if a remote delivery caused a very long error message to be
653 recorded - for instance if somebody sent an entire SpamAssassin report back
654 as a large number of 550 error lines. This bug was coincidentally fixed by
655 increasing the size of one of Exim's internal buffers (big_buffer) that
656 happened as part of the Exiscan merge. However, to be on the safe side, I
657 have made the code more robust (and fixed the comments that describe what
658 is going on).
66053. Now that there can be additional text after "Completed" in log lines (if
661 the queue_time_overall log selector is set), a one-byte patch to exigrep
662 was needed to allow it to recognize "Completed" as not the last thing in
663 the line.
66554. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
666 patch that reportedly fixes this has been added. I am not expert enough to
667 create a test for it. This is what the patch creator wrote:
669 "I found a little strange behaviour of ldap code when working with
670 Windows 2003 AD Domain, where users was placed in more than one
671 Organization Units. When I tried to give exim partial DN, the exit code
672 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
673 But simultaneously result of request was absolutely normal ldap result,
674 so I produce this patch..."
676 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
677 REFERENCE, so I have modified the code to exclude the patch when that macro
678 is not defined.
68055. Some experimental protocols are using DNS PTR records for new purposes. The
681 keys for these records are domain names, not reversed IP addresses. The
682 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
683 leaves it alone. Component reversal etc. now happens only for IP addresses.
ea3a6f44 684 CAN-2005-0021
7102e136 685
68656. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
68857. Double the size of the debug message buffer (to 2048) so that more of very
689 long debug lines gets shown.
69158. The exicyclog utility now does better if the number of log files to keep
692 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
69459. Two changes related to the smtp_active_hostname option:
696 (1) $smtp_active_hostname is now available as a variable.
697 (2) The default for smtp_banner uses $smtp_active_hostname instead
698 of $primary_hostname.
70060. The host_aton() function is supposed to be passed a string that is known
701 to be a valid IP address. However, in the case of IPv6 addresses, it was
702 not checking this. This is a hostage to fortune. Exim now panics and dies
703 if the condition is not met. A case was found where this could be provoked
704 from a dnsdb PTR lookup with an IPv6 address that had more than 8
705 components; fortuitously, this particular loophole had already been fixed
706 by change 4.50/55 above.
708 If there are any other similar loopholes, the new check in host_aton()
709 itself should stop them being exploited. The report I received stated that
710 data on the command line could provoke the exploit when Exim was running as
711 exim, but did not say which command line option was involved. All I could
712 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
713 running as the user.
ea3a6f44 714 CAN-2005-0021
71661. There was a buffer overflow vulnerability in the SPA authentication code
717 (which came originally from the Samba project). I have added a test to the
718 spa_base64_to_bits() function which I hope fixes it.
ea3a6f44 719 CAN-2005-0022
b975ba52 720
72162. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
722 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
72463. The daemon start-up calls getloadavg() while still root for those OS that
725 need the first call to be done as root, but it missed one case: when
726 deliver_queue_load_max is set with deliver_drop_privilege. This is
727 necessary for the benefit of the queue runner, because there is no re-exec
728 when deliver_drop_privilege is set.
73064. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
731 This has been fixed.
73365. Caching of lookup data for "hosts =" ACL conditions, when a named host list
734 was in use, was not putting the data itself into the right store pool;
735 consequently, it could be overwritten for a subsequent message in the same
736 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
737 the caching.)
73966. Added hosts_max_try_hardlimit to the smtp transport, default 50.
74167. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
742 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
743 the function were treating the return as a boolean value, which happened to
744 work because 0=false and not-0=true, but is not correct code.
74668. The host_aton() function was not handling scoped IPv6 addresses (those
747 with, for example, "%eth0" on the end) correctly.
74969. Fixed some compiler warnings in acl.c for the bitmaps specified with
750 negated items (that is, ~something) in unsigned ints. Some compilers
751 apparently mutter when there is no cast.
75370. If an address verification called from an ACL failed, and did not produce a
754 user-specific message (i.e. there was only a "system" message), nothing was
755 put in $acl_verify_message. In this situation, it now puts the system
756 message there.
75871. Change 4.23/11 added synchronization checking at the start of an SMTP
759 session; change 4.31/43 added the unwanted input to the log line - except
760 that it did not do this in the start of session case. It now does.
76272. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
763 This is wrong and can cause the other end to generate a synchronization
764 error if it is another Exim or anything else that does the synchronization
765 check. A QUIT command is no longer sent after a timeout.
76773. $host_lookup_deferred has been added, to make it easier to detect DEFERs
768 during host lookups.
77074. The defer_ok option of callout verification was not working if it was used
771 when verifying addresses in header lines, that is, for this case:
773 verify = header_sender/callout=defer_ok
77575. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
776 those file descriptors could be used for SMTP connections. If anything
777 wrote to stderr (the example that came up was "warn" in embedded Perl), it
778 could be sent to the SMTP client, causing chaos. The daemon now opens
779 stdin, stdout, and stderr to /dev/null when it puts itself into the
780 background.
78276. Arrange for output from Perl's "warn" command to be written to Exim's main
783 log by default. The user can override this with suitable Perl magic.
78577. The use of log_message on a "discard" ACL verb, which is supposed to add to
786 the log message when discard triggers, was not working for the DATA ACL or
787 for the non-SMTP ACL.
78978. Error message wording change in sieve.c.
79179. If smtp_accept_max_per_host was set, the number of connections could be
792 restricted to fewer than expected, because the daemon was trying to set up
793 a new connection before checking whether the processes handling previous
794 connections had finished. The check for completed processes is now done
795 earlier. On busy systems, this bug wouldn't be noticed because something
796 else would have woken the daemon, and it would have reaped the completed
797 process earlier.
79980. If a message was submitted locally by a user whose login name contained one
800 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
801 It caused a spool format error. I have fixed the spool reading code. A
802 related problem was that the "from" clause in the Received: line became
803 illegal because of the space(s). It is now covered by ${quote_local_part.
80581. Included the latest eximstats from Steve (adds average sizes to HTML Top
806 tables).
80882. Updated OS/Makefile-AIX as per message from Mike Meredith.
81083. Patch from Sieve maintainer to fix unterminated string problem in
811 "vacation" handling.
81384. Some minor changes to the Linux configuration files to help with other
814 OS variants using glibc.
81685. One more patch for Sieve to update vacation handling to latest spec.
495ae4b0 818
820See the note above about the 4.44 and 4.50 releases.
824Exim version 4.44
827 1. Change 4.43/35 introduced a bug that caused file counts to be
828 incorrectly computed when quota_filecount was set in an appendfile
829 transport
831 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
832 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
833 place.
835 3. Give more explanation in the error message when the command for a transport
836 filter fails to execute.
838 4. There are several places where Exim runs a non-Exim command in a
839 subprocess. The SIGUSR1 signal should be disabled for these processes. This
840 was being done only for the command run by the queryprogram router. It is
841 now done for all such subprocesses. The other cases are: ${run, transport
842 filters, and the commands run by the lmtp and pipe transports.
844 5. Some older OS have a limit of 256 on the maximum number of file
845 descriptors. Exim was using setrlimit() to set 1000 as a large value
846 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
847 systems. I've change it so that if it can't get 1000, it tries for 256.
849 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
850 was an oversight, and furthermore, ever since the addition of extra
851 controls (e.g. 4.43/32), the checks on when to allow different forms of
852 "control" were broken. There should now be diagnostics for all cases when a
853 control that does not make sense is encountered.
855 7. $recipients is now available in the predata ACL (oversight).
857 8. Tidy the search cache before the fork to do a delivery from a message
858 received from the command line. Otherwise the child will trigger a lookup
859 failure and thereby defer the delivery if it tries to use (for example) a
860 cached ldap connection that the parent has called unbind on.
862 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
863 of $address_data from the recipient verification was clobbered by the
864 sender verification.
86610. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
867 was its contents. (It was OK if the option was not defined at all.)
86911. A "Completed" log line is now written for messages that are removed from
870 the spool by the -Mrm option.
87212. $host_address is now set to the target address during the checking of
873 ignore_target_hosts.
87513. When checking ignore_target_hosts for an ipliteral router, no host name was
876 being passed; this would have caused $sender_host_name to have been used if
877 matching the list had actually called for a host name (not very likely,
878 since this list is usually IP addresses). A host name is now passed as
879 "[x.x.x.x]".
88114. Changed the calls that set up the SIGCHLD handler in the daemon to use the
882 code that specifies a non-restarting handler (typically sigaction() in
883 modern systems) in an attempt to fix a rare and obscure crash bug.
88515. Narrowed the window for a race in the daemon that could cause it to ignore
886 SIGCHLD signals. This is not a major problem, because they are used only to
887 wake it up if nothing else does.
88916. A malformed maildirsize file could cause Exim to calculate negative values
890 for the mailbox size or file count. Odd effects could occur as a result.
891 The maildirsize information is now recalculated if the size or filecount
892 end up negative.
89417. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
895 support for a long time. Removed HAVE_SYS_VFS_H.
ea3a6f44 89718. Updated exipick to current release from John Jetmore.
89919. Allow an empty sender to be matched against a lookup in an address list.
900 Previously the only cases considered were a regular expression, or an
901 empty pattern.
90320. Exim went into a mad DNS lookup loop when doing a callout where the
904 host was specified on the transport, if the DNS lookup yielded more than
905 one IP address.
90721. The RFC2047 encoding function was originally intended for short strings
908 such as real names; it was not keeping to the 75-character limit for
909 encoded words that the RFC imposes. It now respects the limit, and
910 generates multiple encoded words if necessary. To be on the safe side, I
911 have increased the buffer size for the ${rfc2047: expansion operator from
912 1024 to 2048 bytes.
bbe902f0 913
91422. Failure to deliver a bounce message always caused it to be frozen, even if
915 there was an errors_to setting on the router. The errors_to setting is now
916 respected.
91823. If an IPv6 address is given for -bh or -bhc, it is now converted to the
919 canonical form (fully expanded) before being placed in
920 $sender_host_address.
92224. Updated eximstats to version 1.33
92425. Include certificate and key file names in error message when GnuTLS fails
925 to set them up, because the GnuTLS error message doesn't include the name
926 of the failing file when there is a problem reading it.
92826. Expand error message when OpenSSL has problems setting up cert/key files.
ea3a6f44 929 As per change 25.
bbe902f0 930
93127. Reset the locale to "C" after calling embedded Perl, in case it was changed
932 (this can affect the format of dates).
bbe902f0 933
93428. exim_tidydb, when checking for the continued existence of a message for
935 which it has found a message-specific retry record, was not finding
936 messages that were in split spool directories. Consequently, it was
937 deleting retry records that should have stayed in existence.
93929. eximstats updated to version 1.35
940 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
941 1.35 - bugfix such that pie charts by volume are generated correctly
94330. The SPA authentication driver was not abandoning authentication and moving
944 on to the next authenticator when an expansion was forced to fail,
945 contradicting the general specification for all authenticators. Instead it
946 was generating a temporary error. It now behaves as specified.
bbe902f0 947
94831. The default ordering of permitted cipher suites for GnuTLS was pessimal
949 (the order specifies the preference for clients). The order is now AES256,
950 AES128, 3DES, ARCFOUR128.
bbe902f0 951
95231. Small patch to Sieve code - explicitly set From: when generating an
953 autoreply.
bbe902f0 954
95532. Exim crashed if a remote delivery caused a very long error message to be
956 recorded - for instance if somebody sent an entire SpamAssassin report back
957 as a large number of 550 error lines. This bug was coincidentally fixed by
958 increasing the size of one of Exim's internal buffers (big_buffer) that
959 happened as part of the Exiscan merge. However, to be on the safe side, I
960 have made the code more robust (and fixed the comments that describe what
961 is going on).
bbe902f0 962
96333. Some experimental protocols are using DNS PTR records for new purposes. The
964 keys for these records are domain names, not reversed IP addresses. The
965 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
966 leaves it alone. Component reversal etc. now happens only for IP addresses.
967 CAN-2005-0021
96934. The host_aton() function is supposed to be passed a string that is known
970 to be a valid IP address. However, in the case of IPv6 addresses, it was
971 not checking this. This is a hostage to fortune. Exim now panics and dies
972 if the condition is not met. A case was found where this could be provoked
973 from a dnsdb PTR lookup with an IPv6 address that had more than 8
974 components; fortuitously, this particular loophole had already been fixed
975 by change 4.50/55 or 4.44/33 above.
977 If there are any other similar loopholes, the new check in host_aton()
978 itself should stop them being exploited. The report I received stated that
979 data on the command line could provoke the exploit when Exim was running as
980 exim, but did not say which command line option was involved. All I could
981 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
982 running as the user.
983 CAN-2005-0021
98535. There was a buffer overflow vulnerability in the SPA authentication code
986 (which came originally from the Samba project). I have added a test to the
987 spa_base64_to_bits() function which I hope fixes it.
988 CAN-2005-0022
99036. The daemon start-up calls getloadavg() while still root for those OS that
991 need the first call to be done as root, but it missed one case: when
992 deliver_queue_load_max is set with deliver_drop_privilege. This is
993 necessary for the benefit of the queue runner, because there is no re-exec
994 when deliver_drop_privilege is set.
bbe902f0 995
99637. Caching of lookup data for "hosts =" ACL conditions, when a named host list
997 was in use, was not putting the data itself into the right store pool;
998 consequently, it could be overwritten for a subsequent message in the same
999 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1000 the caching.)
bbe902f0 1001
100238. Sometimes the final signoff response after QUIT could fail to get
1003 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1004 < 0 before doing a fflush(). This bug looks as though it goes back to the
1005 introduction of TLS in release 3.20, but "sometimes" must have been rare
1006 because the tests only now provoked it.
1009Exim version 4.43
1012 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1013 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1014 Now it returns an integer. A number of calls were still expecting a T/F
1015 return. Fortuitously, in all cases, the tests worked in OK situations,
1016 which is the norm. However, things would have gone wrong on any write
1017 failures on the smtp file descriptor. This function is used when sending
1018 messages over SMTP and also when doing verify callouts.
1020 2. When Exim is called to do synchronous delivery of a locally submitted
1021 message (the -odf or -odi options), it no longer closes stderr before doing
1022 the delivery.
1024 3. Implemented the mua_wrapper option.
1026 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1028 5. Implemented the functions header_remove(), header_testname(),
1029 header_add_at_position(), and receive_remove_recipient(), and exported them
1030 to local_scan().
1032 6. If an ACL "warn" statement specified the addition of headers, Exim already
1033 inserted X-ACL-Warn: at the start if there was no header name. However, it
1034 was not making this test for the second and subsequent header lines if
1035 there were newlines in the string. This meant that an invalid header could
1036 be inserted if Exim was badly configured.
1038 7. Allow an ACL "warn" statement to add header lines at the start or after all
1039 the Received: headers, as well as at the end.
1041 8. Added the rcpt_4xx retry error code.
1043 9. Added postmaster_mailfrom=xxx to callout verification option.
104510. Added mailfrom=xxxx to the callout verification option, for verify=
1046 header_sender only.
104811. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1049 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
105112. Inserted some casts to stop certain compilers warning when using pointer
1052 differences as field lengths or precisions in printf-type calls (mostly
1053 affecting debugging statements).
105513. Added optional readline() support for -be (dynamically loaded).
105714. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1058 same clock tick as a message's arrival, so that its received time was the
1059 same as the "first fail" time on the retry record, and that message
1060 remained on the queue past the ultimate address timeout, every queue runner
1061 would try a delivery (because it was past the ultimate address timeout) but
1062 after another failure, the ultimate address timeout, which should have then
1063 bounced the address, did not kick in. This was a "< instead of <=" error;
1064 in most cases the first failure would have been in the next clock tick
1065 after the received time, and all would be well.
106715. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1068 being recognized when the domain list was tested by the match_domain
1069 condition in an expansion string.
107116. Added the ${str2b64: operator.
107317. Exim was always calling setrlimit() to set a large limit for the number of
1074 processes, without checking whether the existing limit was already
1075 adequate. (It did check for the limit on file descriptors.) Furthermore,
1076 errors from getrlimit() and setrlimit() were being ignored. Now they are
1077 logged to the main and panic logs, but Exim does carry on, to try to do its
1078 job under whatever limits there are.
108018. Imported PCRE 5.0.
108219. Trivial typo in log message " temporarily refused connection" (the leading
1083 space).
108520. If the log selector return_path_on_delivery was set and an address was
1086 redirected to /dev/null, the delivery process crashed because it assumed
1087 that a return path would always be set for a "successful" delivery. In this
1088 case, the whole delivery is bypassed as an optimization, and therefore no
1089 return path is set.
109121. Internal re-arrangement: the function for sending a challenge and reading
1092 a response while authentication was assuming a zero-terminated challenge
1093 string. It's now changed to take a pointer and a length, to allow for
1094 binary data in such strings.
109622. Added the cyrus_sasl authenticator (code supplied by MBM).
109823. Exim was not respecting finduser_retries when seeking the login of the
1099 uid under which it was called; it was always trying 10 times. (The default
1100 setting of finduser_retries is zero.) Also, it was sleeping after the final
1101 failure, which is pointless.
110324. Implemented tls_on_connect_ports.
110525. Implemented acl_smtp_predata.
110726. If the domain in control=submission is set empty, Exim assumes that the
1108 authenticated id is a complete email address when it generates From: or
1109 Sender: header lines.
111127. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1112 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1113 chown and chgrp in /bin and hostname in /usr/bin.
111528. Exim was keeping the "process log" file open after each use, just as it
1116 does for the main log. This opens the possibility of it remaining open for
1117 long periods when the USR1 signal hits a daemon. Occasional processlog
1118 errors were reported, that could have been caused by this. Anyway, it seems
1119 much more sensible not to leave this file open at all, so that is what now
1120 happens.
112229. The long-running daemon process does not normally write to the log once it
1123 has entered its main loop, and it closes the log before doing so. This is
1124 so that log files can straightforwardly be renamed and moved. However,
1125 there are a couple of unusual error situations where the daemon does write
1126 log entries, and I had neglected to close the log afterwards.
112830. The text of an SMTP error response that was received during a remote
1129 delivery was being truncated at 512 bytes. This is too short for some of
1130 the long messages that one sometimes sees. I've increased the limit to
1131 1024.
113331. It is now possible to make retry rules that apply only when a message has a
1134 specific sender, in particular, an empty sender.
113632. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1137 it possible to be selective about when SMTP synchronization is enforced.
113933. Added "control = caseful_local_part" and "control = "caselower_local_part".
114132. Implemented hosts_connection_nolog.
114333. Added an ACL for QUIT.
114534. Setting "delay_warning=" to disable warnings was not working; it gave a
1146 syntax error.
114835. Added mailbox_size and mailbox_filecount to appendfile.
115036. Added control = no_multiline_responses to ACLs.
115237. There was a bug in the logic of the code that waits for the clock to tick
1153 in the case where the clock went backwards by a substantial amount such
1154 that the microsecond fraction of "now" was more than the microsecond
1155 fraction of "then" (but the whole seconds number was less).
115738. Added support for the libradius Radius client library this is found on
1158 FreeBSD (previously only the radiusclient library was supported).
1161Exim version 4.42
1164 1. When certain lookups returned multiple values in the form name=value, the
1165 quoting of the values was not always being done properly. Specifically:
1166 (a) If the value started with a double quote, but contained no whitespace,
1167 it was not quoted.
1168 (b) If the value contained whitespace other than a space character (i.e.
1169 tabs or newlines or carriage returns) it was not quoted.
1170 This fix has been applied to the mysql and pgsql lookups by writing a
1171 separate quoting function and calling it from the lookup code. The fix
1172 should probably also be applied to nisplus, ibase and oracle lookups, but
1173 since I cannot test any of those, I have not disturbed their existing code.
1175 2. A hit in the callout cache for a specific address caused a log line with no
1176 reason for rejecting RCPT. Now it says "Previous (cached) callout
1177 verification failure".
1179 3. There was an off-by-one bug in the queryprogram router. An over-long
1180 return line was truncated at 256 instead of 255 characters, thereby
1181 overflowing its buffer with the terminating zero. As well as fixing this, I
1182 have increased the buffer size to 1024 (and made a note to document this).
1184 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1185 when Exim is waiting for an SMTP response from a remote server, Exim
1186 restarts its select() call on the socket, thereby resetting its timeout.
1187 This is not a problem when such interrupts are rare. Somebody set up a cron
1188 job to run exiwhat every 2 minutes, which is less than the normal select()
1189 timeout (5 or 10 minutes). This meant that the select() timeout never
1190 kicked in because it was always reset. I have fixed this by comparing the
1191 time when an interrupt arrives with the time at the start of the first call
1192 to select(). If more time than the timeout has elapsed, the interrupt is
1193 treated as a timeout.
1195 5. Some internal re-factoring in preparation for the addition of Sieve
1196 extensions (by MH). In particular, the "personal" test is moved to a
1197 separate function, and given an option for scanning Cc: and Bcc: (which is
1198 not set for Exim filters).
1200 6. When Exim created an email address using the login of the caller as the
1201 local part (e.g. when creating a From: or Sender: header line), it was not
1202 quoting the local part when it contained special characters such as @.
1204 7. Installed new OpenBSD configuration files.
1206 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1207 try to make them clearer.
1209 9. Callout options, other than the timeout value, were being ignored when
1210 verifying sender addresses in header lines. For example, when using
1212 verify = header_sender/callout=no_cache
1214 the cache was (incorrectly) being used.
121610. Added a missing instance of ${EXE} to the exim_install script; this affects
1217 only the Cygwin environment.
121911. When return_path_on_delivery was set as a log selector, if different remote
1220 addresses in the same message used different return paths and parallel
1221 remote delivery occurred, the wrong values would sometimes be logged.
1222 (Whenever a remote delivery process finished, the return path value from
1223 the most recently started remote delivery process was logged.)
122512. RFC 3848 specifies standard names for the "with" phrase in Received: header
1226 lines when AUTH and/or TLS are in use. This is the "received protocol"
1227 field. Exim used to use "asmtp" for authenticated SMTP, without any
1228 indication (in the protocol name) for TLS use. Now it follows the RFC and
1229 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1230 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1231 names appear in log lines as well as in Received: header lines.
123313. Installed MH's patches for Sieve to add the "copy" and "vacation"
1234 extensions, and comparison tests, and to fix some bugs.
123614. Changes to the "personal" filter test:
1238 (1) The test was buggy in that it was just doing the equivalent of
1239 "contains" tests on header lines. For example, if a user's address was
1240 anne@some.where, the "personal" test would incorrectly be true for
1242 To: susanne@some.where
1244 This test is now done by extracting each address from the header in turn,
1245 and checking the entire address. Other tests that are part of "personal"
1246 are now done using regular expressions (for example, to check local parts
1247 of addresses in From: header lines).
1249 (2) The list of non-personal local parts in From: addresses has been
1250 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1251 taken from the Sieve specification recommendations.
1253 (3) If the message contains any header line starting with "List-" it is
1254 treated as non-personal.
1256 (4) The test for "circular" in the Subject: header line has been removed
1257 because it now seems ill-conceived.
125915. Minor typos in src/EDITME comments corrected.
126116. Installed latest exipick from John Jetmore.
126317. If headers_add on a router specified a text string that was too long for
1264 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1265 of string_sprintf() is now avoided.
126718. $message_body_size was not set (it was always zero) when running the DATA
1268 ACL and the local_scan() function.
127019. For the "mail" command in an Exim filter, no default was being set for
1271 the once_repeat time, causing a random time value to be used if "once" was
1272 specified. (If the value happened to be <= 0, no repeat happened.) The
1273 default is now 0s, meaning "never repeat". The "vacation" command was OK
1274 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1275 (I found it when inspecting the code).
127720. There is now an overall timeout for performing a callout verification. It
1278 defaults to 4 times the callout timeout, which applies to individual SMTP
1279 commands during the callout. The overall timeout applies when there is more
1280 than one host that can be tried. The timeout is checked before trying the
1281 next host. This prevents very long delays if there are a large number of
1282 hosts and all are timing out (e.g. when the network connections are timing
1283 out). The value of the overall timeout can be changed by specifying an
1284 additional sub-option for "callout", called "maxwait". For example:
1286 verify = sender/callout=5s,maxwait=20s
128821. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1289 to the end before writing, but this should make it even safer).
129122. Exim was forgetting that it had advertised PIPELINING for the second and
1292 subsequent messages on an SMTP connection. It was also not resetting its
1293 memory on STARTTLS and an internal HELO.
129523. When Exim logs an SMTP synchronization error within a session, it now
1296 records whether PIPELINING has been advertised or not.
129824. Added 3 instances of "(long int)" casts to time_t variables that were being
1299 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1300 rather than long int.
130225. Installed the latest Cygwin configuration files from the Cygwin maintainer.
130426. Added the never_mail option to autoreply.
1307Exim version 4.41
1310 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1311 crash if the getsockname() call failed; this can happen if a connection is
1312 closed very soon after it is established. The problem was simply in the
1313 order in which certain operations were done, causing Exim to try to write
1314 to the SMTP stream before it had set up the file descriptor. The bug has
1315 been fixed by making things happen in the correct order.
1318Exim version 4.40
1321 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1322 before the connection was closed, thus losing the rejection response.
1324 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1325 some early Solaris releases, but causes trouble in current releases where
1326 socklen_t is defined.
1328 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1329 always exist.
1331 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1332 configured.
1334 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1335 that releasing the top of it at the end releases what was used for sub-
1336 expansions (unless the block got too big). However, discard this block if
1337 the first thing is a variable or header, so that we can use its block when
1338 it is dynamic (useful for very large $message_headers, for example).
1340 6. Lookups now cache *every* query, not just the most recent. A new, separate
1341 store pool is used for this. It can be recovered when all lookup caches are
1342 flushed. Lookups now release memory at the end of their result strings.
1343 This has involved some general refactoring of the lookup sources.
1345 7. Some code has been added to the store_xxx() functions to reduce the amount
1346 of flapping under certain conditions.
1348 8. log_incoming_interface used to affect only the <= reception log lines. Now
1349 it causes the local interface and port to be added to several more SMTP log
1350 lines, for example "SMTP connection from", and rejection lines.
1352 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
135410. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
135611. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1357 could be overwritten at the end of the current message (or the start of a
1358 new message if it was set in a HELO ACL). The value is now preserved for
1359 the duration of the SMTP connection.
136112. If a transport had a headers_rewrite setting, and a matching header line
1362 contained an unqualified address, that address was qualified, even if it
1363 did not match any rewriting rules. The underlying bug was that the values
1364 of the flags that permit the existence of unqualified sender and recipient
1365 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1366 non-local messages, and by -bnq for local messages) were not being
1367 preserved with the message after it was received.
136913. When Exim was logging an SMTP synchronization error, it could sometimes log
1370 "next input=" as part of the text comprising the host identity instead of
1371 the correct text. The code was using the same buffer for two different
1372 strings. However, depending on which order the printing function evaluated
1373 its arguments, the bug did not always show up. Under Linux, for example, my
1374 test suite worked just fine.
137614. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1377 doesn't work with some older versions of Perl. It has been changed to "my",
1378 which in any case is probably the better facility to use.
138015. A really picky compiler found some instances of statements for creating
1381 error messages that either had too many or two few arguments for the format
1382 string.
138416. The size of the buffer for calls to the DNS resolver has been increased
1385 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1386 for addresses that have a lot of PTR records. This alleviates a problem; it
1387 does not fully solve it.
138917. A dnsdb lookup for PTR records that receives more data than will fit in the
1390 buffer now truncates the list and logs the incident, which is the same
1391 action as happens when Exim is looking up a host name and its aliases.
1392 Previously in this situation something unpredictable would happen;
1393 sometimes it was "internal error: store_reset failed".
139518. If a server dropped the connection unexpectedly when an Exim client was
1396 using GnuTLS and trying to read a response, the client delivery process
1397 crashed while trying to generate an error log message.
139919. If a "warn" verb in an ACL added multiple headers to a message in a single
1400 string, for example:
1402 warn message = H1: something\nH2: something
1404 the text was added as a single header line from Exim's point of view
1405 though it ended up OK in the delivered message. However, searching for the
1406 second and subsequent header lines using $h_h2: did not work. This has been
1407 fixed. Similarly, if a system filter added multiple headers in this way,
1408 the routers could not see them.
141020. Expanded the error message when iplsearch is called with an invalid key to
1411 suggest using net-iplsearch in a host list.
141321. When running tests using -bh, any delays imposed by "delay" modifiers in
1414 ACLs are no longer actually imposed (and a message to that effect is
1415 output).
141722. If a "gecos" field in a passwd entry contained escaped characters, in
1418 particular, if it contained a \" sequence, Exim got it wrong when building
1419 a From: or a Sender: header from that name. A second bug also caused
1420 incorrect handling when an unquoted " was present following a character
1421 that needed quoting.
142323. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1424 was not being matched caselessly.
142624. Arranged for all hyphens in the exim.8 source to be escaped with
1427 backslashes.
142925. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1430 properly. Recipient callout cache records were still being keyed to include
1431 the sender, even when use_sender was set false. This led to far more
1432 callouts that were necessary. The sender is no longer included in the key
1433 when use_sender is false.
143526. Added "control = submission" modifier to ACLs.
143727. Added the ${base62d: operator to decode base 62 numbers.
143928. dnsdb lookups can now access SRV records.
144129. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1442 the configuration file.
144430. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1445 (-v) mode. This makes the output for a verbose queue run more intelligible.
144731. Added a use_postmaster feature to recipient callouts.
144932. Added the $body_zerocount variable, containing the number of binary zero
1450 bytes in the message body.
145233. The time of last modification of the "new" subdirectory is now used as the
1453 "mailbox time last read" when there is a quota error for a maildir
1454 delivery.
145634. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
145835. Added +ignore_unknown as a special item in host lists.
146036. Code for decoding IPv6 addresses in host lists is now included, even if
1461 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1462 address was recognized as an IP address, but was then not correctly decoded
1463 into binary, causing unexpected and incorrect effects when compared with
1464 another IP address.
1467Exim version 4.34
1470 1. Very minor rewording of debugging text in manualroute to say "list of
1471 hosts" instead of "hostlist".
1473 2. If verify=header_syntax was set, and a header line with an unqualified
1474 address (no domain) and a large number of spaces between the end of the
1475 name and the colon was received, the reception process suffered a buffer
1476 overflow, and (when I tested it) crashed. This was caused by some obsolete
1477 code that should have been removed. The fix is to remove it!
1479 3. When running in the test harness, delay a bit after writing a bounce
1480 message to get a bit more predictability in the log output.
1482 4. Added a call to search_tidyup() just before forking a reception process. In
1483 theory, someone could use a lookup in the expansion of smtp_accept_max_
1484 per_host which, without the tidyup, could leave open a database connection.
1486 5. Added the variables $recipient_data and $sender_data which get set from a
1487 lookup success in an ACL "recipients" or "senders" condition, or a router
1488 "senders" option, similar to $domain_data and $local_part_data.
1490 6. Moved the writing of debug_print from before to after the "senders" test
1491 for routers.
1493 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1494 problems for message scanning, either using a data ACL, or using
1495 local_scan() because the Received: header was not generated till after they
1496 were called (in order to set the time as the time of reception completion).
1497 I have revised the way this works. The header is now generated after the
1498 body is received, but before the ACL or local_scan() are called. After they
1499 are run, the timestamp in the header is updated.
1502Exim version 4.33
1505 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1506 before starting a queue runner without re-exec. This happened only when
1507 deliver_drop_privilege was set or when the Exim user was set to root. The
1508 effect of the bug was that timeouts during subsequent deliveries caused
1509 crashes instead of being properly handled. The handler is now left at its
1510 default (and expected) setting.
1512 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1513 message, again when deliver_drop_privilege is set or Exim is run as root.
1514 The bug described in (1) was not present in this case, but the tidying up
1515 of the other signals was missing. I have made the two cases consistent.
1517 3. The ignore_target_hosts setting on a manualroute router was being ignored
1518 for hosts that were looked up using the /MX notation.
1520 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1521 in domain lists.
1523 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1524 operated on the sender address. After changing the $sender_address to <>
1525 for the sender address verify, Exim was re-instated it as the original
1526 (before rewriting) address, but remembering that it had rewritten it, so it
1527 wasn't rewriting it again. This bug also had the effect of breaking the
1528 sender address verification caching when the sender address was rewritten.
1530 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1531 This has been changed so that if the ip literal address matches
1532 ignore_target_hosts, the router declines.
1534 7. Added expansion conditions match_domain, match_address, and match_local_
1535 part (NOT match_host).
1537 8. The placeholder for the Received: header didn't have a length field set.
1539 9. Added code to Exim itself and to exim_lock to test for a specific race
1540 condition that could lead to file corruption when using MBX delivery. The
1541 issue is with the lockfile that is created in /tmp. If this file is removed
1542 after a process has opened it but before that process has acquired a lock,
1543 there is the potential for a second process to recreate the file and also
1544 acquire a lock. This could lead to two Exim processes writing to the file
1545 at the same time. The added code performs the same test as UW imapd; it
1546 checks after acquiring the lock that its file descriptor still refers to
1547 the same named file.
154910. The buffer for building added header lines was of fixed size, 8192 bytes.
1550 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1551 when Exim is built.
155311. Added the smtp_active_hostname option. If used, this will typically be made
1554 to depend on the incoming interface address. Because $interface_address is
1555 not set up until the daemon has forked a reception process, error responses
1556 that can happen earlier (such as "too many connections") no longer contain
1557 a host name.
155912. If an expansion in a condition on a "warn" statement fails because a lookup
1560 defers, the "warn" statement is abandoned, and the next ACL statement is
1561 processed. Previously this caused the whole ACL to be aborted.
156313. Added the iplsearch lookup type.
156514. Added ident_timeout as a log selector.
156715. Added tls_certificate_verified as a log selector.
156916. Added a global option tls_require_ciphers (compare the smtp transport
1570 option of the same name). This controls incoming TLS connections.
157217. I finally figured out how to make tls_require_ciphers do a similar thing
1573 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1574 before starting the TLS session.
157618. Tabs are now shown as \t in -bP output.
157819. If the log selector return_path_on_delivery was set, Exim crashed when
1579 bouncing a message because it had too many Received: header lines.
158120. If two routers both had headers_remove settings, and the first one included
1582 a superfluous trailing colon, the final name in the first list and the
1583 first name in the second list were incorrectly joined into one item (with a
1584 colon in the middle).
1587Exim version 4.32
1590 1. Added -C and -D options to the exinext utility, mainly to make it easier
1591 to include in the automated testing, but these could be helpful when
1592 multiple configurations are in use.
1594 2. The exinext utility was not formatting the output nicely when there was
1595 an alternate port involved in the retry record key, nor when there was a
1596 message id as well (for retries that were specific to a specific message
1597 and a specific host). It was also confused by IPv6 addresses, because of
1598 the additional colons they contain. I have fixed the IPv4 problem, and
1599 patched it up to do a reasonable job for IPv6.
1601 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1602 delivery, the log line now contains "pipelined" if PIPELINING was used.
1604 4. An SMTP transport process used to panic and die if the bind() call to set
1605 an explicit outgoing interface failed. This has been changed; it is now
1606 treated in the same way as a connect() failure.
1608 5. A reference to $sender_host_name in the part of a conditional expansion
1609 that was being skipped was still causing a DNS lookup. This no longer
1610 occurs.
1612 6. The def: expansion condition was not recognizing references to header lines
1613 that used bh_ and bheader_.
1615 7. Added the _cache feature to named lists.
1617 8. The code for checking quota_filecount in the appendfile transport was
1618 allowing one more file than it should have been.
1620 9. For compatibility with Sendmail, the command line option
1622 -prval:sval
1624 is equivalent to
1626 -oMr rval -oMs sval
1628 and sets the incoming protocol and host name (for trusted callers). The
1629 host name and its colon can be omitted when only the protocol is to be set.
1630 Note the Exim already has two private options, -pd and -ps, that refer to
1631 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1632 "s", but I don't think that's a major issue.
163410. A number of refactoring changes to the code, none of which should affect
1635 Exim's behaviour:
1637 (a) The number of logging options was getting close to filling up the
1638 32-bit word that was used as a bit map. I have split them into two classes:
1639 those that are passed in the argument to log_write(), and those that are
1640 only ever tested independently outside of that function. These are now in
1641 separate 32-bit words, so there is plenty of room for expansion again.
1642 There is no change in the user interface or the logging behaviour.
1644 (b) When building, for example, log lines, the code previously used a
1645 macro that called string_cat() twice, in order to add two strings. This is
1646 not really sufficiently general. Furthermore, there was one instance where
1647 it was actually wrong because one of the argument was used twice, and in
1648 one call a function was used. (As it happened, calling the function twice
1649 did not affect the overall behaviour.) The macro has been replaced by a
1650 function that can join an arbitrary number of extra strings onto a growing
1651 string.
1653 (c) The code for expansion conditions now uses a table and a binary chop
1654 instead of a serial search (which was left over from when there were very
1655 few conditions). Also, it now recognizes conditions like "pam" even when
1656 the relevant support is not compiled in: a suitably worded error message is
1657 given if an attempt is made to use such a condition.
165911. Added ${time_interval:xxxxx}.
166112. A bug was causing one of the ddress fields not to be passed back correctly
1662 from remote delivery subprocesses. The field in question was not being
1663 subsequently used, so this caused to problems in practice.
166513. Added new log selectors queue_time and deliver_time.
166714. Might have fixed a bug in maildirsizefile handling that threw up
1668 "unexpected character" debug warnings, and recalculated the data
1669 unnecessarily. In any case, I expanded the warning message to give more
1670 information.
167215. Added the message "Restricted characters in address" to the statements in
1673 the default ACL that block characters like @ and % in local parts.
167516. Change 71 for release 4.31 proved to be much less benign that I imagined.
1676 Three changes have been made:
1678 (a) There was a serious bug; a negative response to MAIL caused the whole
1679 recipient domain to be cached as invalid, thereby blocking all messages
1680 to all local parts at the same domain, from all senders. This bug has
1681 been fixed. The domain is no longer cached after a negative response to
1682 MAIL if the sender used is not empty.
1684 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1685 been restored.
1687 (c) A new callout option, "use_sender" has been added for people who want
1688 the modified behaviour.
1691Exim version 4.31
1694 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1695 Larry Rosenman.
1697 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1698 indeed breaks things for older releases.
1700 3. Added additional logging to the case where there is a problem reading data
1701 from a filter that is running in a subprocess using a pipe, in order to
1702 try to track down a specific problem.
1704 4. Testing facility fudge: when running in the test harness and attempting
1705 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1706 getting "No route to host". Convert this to a timeout.
1708 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1709 warning.
1711 6. Some OS don't have socklen_t but use size_t instead. This affects the
1712 fifth argument of getsockopt() amongst other things. This is now
1713 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1714 can be set for individual OS. I have set it for SunOS5, OSF1, and
1715 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1716 some earlier ones do not.
1718 7. Change 4.30/15 was not doing the test caselessly.
1720 8. The standard form for an IPv6 address literal was being rejected by address
1721 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1722 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1723 this, as well as the form without the "IPv6" on the front (but only when
1724 address literals are enabled, of course).
1726 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
172810. Exim crashed if a message with an empty sender address specified by -f
1729 encountered a router with an errors_to setting. This could be provoked only
1730 by a command such as
1732 exim -f "" ...
1734 where an empty string was supplied; "<>" did not hit this bug.
173611. Installed PCRE release 4.5.
173812. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1739 remained set. It is now erased.
174113. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1742 times from message ids (which are base 36 rather than the normal 62).
174414. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1745 were being counted as actual protocol errors, and logged if the log
1746 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1747 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1748 and DATA following a set of rejected RCPTs do not count as protocol errors.
1749 In other words, Exim assumes they were pipelined, though this may not
1750 actually be the case. Of course, in all cases the client gets an
1751 appropriate error code.
175315. If a lookup fails in an ACL condition, a message about the failure may
1754 be available; it is used if testing the ACL cannot continue, because most
1755 such messages specify what the cause of the deferral is. However, some
1756 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1757 that caused an old message to be retained and used if a later statement
1758 caused a defer, replacing the real cause of the deferral.
176016. If an IP address had so many PTR records that the DNS lookup buffer
1761 was not large enough to hold them, Exim could crash while trying to process
1762 the truncated data. It now detects and logs this case.
176417. Further to 4.21/58, another change has been made: if (and only if) the
1765 first line of a message (the first header line) ends with CRLF, a bare LF
1766 in a subsequent header line has a space inserted after it, so as not to
1767 terminate the header.
176918. Refactoring: tidied an ugly bit of code in appendfile that copied data
1770 unnecessarily, used atoi() instead of strtol(), and didn't check the
1771 termination when getting file sizes from file names by regex.
177319. Completely re-implemented the support for maildirsize files, in the light
1774 of a number of problems with the previous contributed implementation
1775 (4.30/29). In particular:
1777 . If the quota is zero, the maildirsize file is maintained, but no quota is
1778 imposed.
1780 . If the maildir directory does not exist, it is created before any attempt
1781 to write a maildirsize file.
1783 . The quota value in the file is just a cache; if the quota is changed in
1784 the transport, the new value overrides.
1786 . A regular expression is available for excluding directories from the
1787 count.
178920. The autoreply transport checks the characters in options that define the
1790 message's headers; it allows continued headers, but it was checking with
1791 isspace() after an embedded newline instead of explicitly looking for a
1792 space or a tab.
179421. If all the "regular" hosts to which an address was routed had passed their
1795 expiry times, and had not reached their retry times, the address was
1796 bounced, even if fallback hosts were defined. Now Exim should go on to try
1797 the fallback hosts.
179922. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1800 equivalent code in the SMTP transport. Some hosts send humungous responses
1801 to HELO/EHLO, more than 1024 it seems.
180323. Refactoring: code in filter.c used (void *) for "any old type" but this
1804 gives compiler warnings in some environments. I've now done it "properly",
1805 using a union.
180724. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1808 (because of problems with the built-in one) was declared to return uschar *
1809 instead of char *, causing compiler failure.
181125. Fixed a file descriptor leak when processing alias/forward files.
181326. Fixed a minor format string issue in dbfn.c.
181527. Typo in exim.c: ("dmbnz" for "dbmnz").
181728. If a filter file refered to $h_xxx or $message_headers, and the headers
1818 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1819 become corrupted.
182129. When a sender address is verified, it is cached, to save repeating the test
1822 when there is more than one recipient in a message. However, when the
1823 verification involves a callout, it is possible for different callout
1824 options to be set for different recipients. It is too complicated to keep
1825 track of this in the cache, so now Exim always runs a verification when a
1826 callout is required, relying on the callout cache for the optimization.
1827 The overhead is duplication of the address routing, but this should not be
1828 too great.
183030. Fixed a bug in callout caching. If a RCPT command caused the sender address
1831 to be verified with callout=postmaster, and the main callout worked but the
1832 postmaster check failed, the verification correctly failed. However, if a
1833 subsequent RCPT command asked for sender verification *without* the
1834 postmaster check, incorrect caching caused this verification also to fail,
1835 incorrectly.
183731. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1838 it was not caching the DNS options (qualify_single, search_parents) that
1839 were used when the lookup failed. A subsequent lookup with different
1840 options therefore always gave the same answer, though there were cases
1841 where it should not have. (Example: a "domains = !$mx_any" option on a
1842 dnslookup router: the "domains" option is always processed without any
1843 widening, but the router might have qualify_single set.) Now Exim uses the
1844 cached value only when the same options are set.
184632. Added John Jetmore's "exipick" utility to the distribution.
184833. GnuTLS: When an attempt to start a TLS session fails for any reason other
1849 than a timeout (e.g. a certificate is required, and is not provided), an
1850 Exim server now closes the connection immediately. Previously it waited for
1851 the client to close - but if the client is SSL, it seems that they each
1852 wait for each other, leading to a delay before one of them times out.
185434: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1855 maintained 0.8.x compatibility because I don't think many are using it, and
1856 it is clearly obsolete.
185835. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1859 transport.
186136. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1862 client certificate was expired. A simple patch fixes this, though I don't
1863 understand the full logic of why the verify callback is called multiple
1864 times.
186637. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1867 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1868 which causes problems with some clients (such as the Certicom SSL Plus
1869 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1870 disables the coutermeasure allowing Eudora to connect."
187238. Exim was not checking that a write() to a log file succeeded. This could
1873 lead to Bad Things if a log got too big, in particular if it hit a file
1874 size limit. Exim now panics and dies if it cannot write to a log file, just
1875 as it does if it cannot open a log file.
187739. Modified OS/Makefile-Linux so that it now contains
1881 The two -D definitions ensure that Exim is compiled with large file
1882 support, which makes it possible to handle log files that are bigger than
1883 2^31.
188540. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
1886 instance) a domain was checked against a named list that involved a lookup,
1887 causing $domain_data to be set, then another domain was checked against the
1888 same list, then the first domain was re-checked, the value of $domain_data
1889 after the final check could be wrong. In particular, if the second check
1890 failed, it could be set empty. This bug probably also applied to
1891 $localpart_data.
189341. The strip_trailing_dot option was not being applied to the address given
1894 with the -f command-line option.
189642. The code for reading a message's header from the spool was incrementing
1897 $received_count, but never initializing it. This meant that the value was
1898 incorrect (doubled) while delivering a message in the same process in which
1899 it was received. In the most common configuration of Exim, this never
1900 happens - a fresh exec is done - but it can happen when
1901 deliver_drop_privilege is set.
190343. When Exim logs an SMTP synchronization error - client data sent too soon -
1904 it now includes up to 150 characters of the unexpected data in the log
1905 line.
190744. The exim_dbmbuild utility uses fixed size buffers for reading input lines
1908 and building data strings. The size of both of these buffers was 10 000
1909 bytes - far larger than anybody would *ever* want, thought I. Needless to
1910 say, somebody hit the limit. I have increased the maximum line length to
1911 20 000 and the maximum data length of concatenated lines to 100 000. I have
1912 also fixed two bugs, because there was no checking on these buffers. Tsk,
1913 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
1914 buffer is too small.
191645. The exim_dbmbuild utility did not support quoted keys, as Exim does in
1917 lsearch lookups. Now it does.
191946. When parsing a route_list item in a manualroute router, a fixed-length
1920 buffer was used for the list of hosts. I made this 1024 bytes long,
1921 thinking that nobody would ever have a list of hosts that long. Wrong.
1922 Somebody had a whole pile of complicated expansion conditions, and the
1923 string was silently truncated, leading to an expansion error. It turns out
1924 that it is easier to change to an unlimited length (owing to other changes
1925 that have happened since this code was originally written) than to build
1926 structure for giving a limitation error. The length of the item that
1927 expands into the list of hosts is now unlimited.
192947. The lsearch lookup could not handle data where the length of text line was
1930 more than 4095 characters. Such lines were truncated, leading to shortened
1931 data being returned. It should now handle lines of any length.
193348. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
1934 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
1935 ACL").
193749. Cosmetic tidy to scripts like exicyclog that are generated by globally
1938 replacing strings such as BIN_DIRECTORY in a source file: the replacement
1939 no longer happens in comment lines. A list of replacements is now placed
1940 at the head of all of the source files, except those whose only change is
1941 to replace PERL_COMMAND in the very first #! line.
194350. Replaced the slow insertion sort in queue.c, for sorting the list of
1944 messages on the queue, with a bottom-up merge sort, using code contributed
1945 by Michael Haardt. This should make operations like -bp somewhat faster on
1946 large queues. It won't affect queue runners, except when queue_run_in_order
1947 is set.
194951. Installed eximstats 1.31 in the distribution.
195152. Added support for SRV lookups to the dnslookup router.
195353. If an ACL referred to $message_body or $message_body_end, the value was not
1954 reset for any messages that followed in the same SMTP session.
195654. The store-handling optimization for building very long strings was not
1957 differentiating between the different store pools. I don't think this
1958 actually made any difference in practice, but I've tidied it.
196055. While running the routers to verify a sender address, $sender_address
1961 was still set to the sender address. This is wrong, because when routing to
1962 send a bounce to the sender, it would be empty. Therefore, I have changed
1963 it so that, while verifying a sender address, $sender_address is set to <>.
1964 (There is no change to what happens when verifying a recipient address.)
196656. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
1967 target A or AAAA records (if not already returned) without resetting the
1968 qualify_single or search_parents options of the DNS resolver. These are
1969 inappropriate in this case because the targets of MX and SRV records must
1970 be FQDNs. A broken DNS record could cause trouble if it happened to have a
1971 target that, when qualified, matched something in the local domain. These
1972 two options are now turned off when doing these lookups.
197457. It seems that at least some releases of Reiserfs (which does not have the
1975 concept of a fixed number of inodes) returns zero and not -1 for the
1976 number of available inodes. This interacted badly with check_spool_inodes,
1977 which assumed that -1 was the "no such thing" setting. What I have done is
1978 to check that the total number of inodes is greater than zero before doing
1979 the test of how many are available.
198158. When a "warn" ACL statement has a log_message modifier, the message is
1982 remembered, and not repeated. This is to avoid a lot of repetition when a
1983 message has many recipients that cause the same warning to be written.
1984 Howewer, Exim was preserving the list of already written lines for an
1985 entire SMTP session, which doesn't seem right. The memory is now reset if a
1986 new message is started.
198859. The "rewrite" debugging flag was not showing the result of rewriting in the
1989 debugging output unless log_rewrite was also set.
199160. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
1992 of (int)(handle) when we know that handle contains (void *)(-1).
199461. The Exim daemon panic-logs an error return when it closes the incoming
1995 connection. However "connection reset by peer" seems to be common, and
1996 isn't really an error worthy of noting specially, so that particular error
1997 is no long logged.
199962. When Exim is trying to find all the local interfaces, it used to panic and
2000 die if the ioctl to get the interface flags failed. However, it seems that
2001 on at least one OS (Solaris 9) it is possible to have an interface that is
2002 included in the list of interfaces, but for which you get a failure error
2003 for this call. This happens when the interface is not "plumbed" into a
2004 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2005 failure of the "get flags" call assumes that the interface is down.
200763. Added a ${eval10: operator, which assumes all numbers are decimal. This
2008 makes life easier for people who are doing arithmetic on fields extracted
2009 from dates, where you often get leading zeros that should not be
2010 interpreted as octal.
201264. Added qualify_domain to the redirect router, to override the global
2013 setting.
201565. If a pathologically long header line contained very many addresses (the
2016 report of this problem mentioned 10 000) and each of them was rewritten,
2017 Exim could use up a very large amount of memory. (It kept on making new
2018 copies of the header line as it rewrote, and never released the old ones.)
2019 At the expense of a bit more processing, the header rewriting function has
2020 been changed so that it no longer eats memory in this way.
202266. The generation of the Received: header has been moved from the time that a
2023 message starts to be received, to the time that it finishes. The timestamp
2024 in the Received: header should now be very close to that of the <= log
2025 line. There are two side-effects of this change:
2027 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2028 logged header lines no longer include the local Received: line, because
2029 it has not yet been created. The same applies to a copy of the message
2030 that is returned to a non-SMTP sender when a message is rejected.
2032 (b) When a filter file is tested using -bf, no additional Received: header
2033 is added to the test message. After some thought, I decided that this
2034 is a bug fix.
2036 This change does not affect the value of $received_for. It is still set
2037 after address rewriting, but before local_scan() is called.
203967. Installed the latest Cygwin-specific files from the Cygwin maintainer.
204168. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2042 gave an unhelpful panic error message, and a defer error. I have managed to
2043 change this behaviour so that it now rejects any supplied certificate,
2044 which seems right, as the list of acceptable certificates is empty.
204669. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2047 gave an unhelpful defer error. I have not managed to make this reject any
2048 supplied certificates, but the error message it gives is "no certificate
2049 supplied", which is not helpful.
205170. exigrep's output now also includes lines that are not associated with any
2052 message, but which match the given pattern. Implemented by a patch from
2053 Martin Sluka, which also tidied up the Perl a bit.
205571. Recipient callout verification, like sender verification, was using <> in
2056 the MAIL FROM command. This isn't really the right thing, since the actual
2057 sender may affect whether the remote host accepts the recipient or not. I
2058 have changed it to use the actual sender in the callout; this means that
2059 the cache record is now keyed on a recipient/sender pair, not just the
2060 recipient address. There doesn't seem to be a real danger of callout loops,
2061 since a callout by the remote host to check the sender would use <>.
2062 [SEE ABOVE: changed after hitting problems.]
206472. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2065 temporary errors. However, in the case of such a code being given after
2066 the end of a data transmission (i.e. after ".") Exim was failing to write
2067 a retry record for the message. (Yes, there was some broken host that was
2068 actually sending 8xx at this point.)
207073. An unknown lookup type in a host list could cause Exim to panic-die when
2071 the list was checked. (An example that provoked this was putting <; in the
2072 middle of a list instead of at the start.) If this happened during a DATA
2073 ACL check, a -D file could be left lying around. This kind of configuration
2074 error no longer causes Exim to die; instead it causes a defer errror. The
2075 incident is still logged to the main and panic logs.
207774. Buglet left over from Exim 3 conversion. The message "too many messages
2078 in one connection" was written to the rejectlog but not the mainlog, except
2079 when address rewriting (yes!) was being logged.
208175. Added write_rejectlog option.
208376. When a system filter was run not as root (that is, when system_filter_user
2084 was set), the values of the $n variables were not being returned to the
2085 main process; thus, they were not subsequently available in the $sn
2086 variables.
208877. Added +return_path_on_delivery log selector.
209078. A connection timeout was being treated differently from recipients deferred
2091 when testing hosts_max_try with a message that was older than the host's
2092 retry timeout. (The host should not be counted, thus allowing all hosts to
2093 be tried at least once before bouncing.) This may have been the cause of an
2094 occasionally reported bug whereby a message would remain on the queue
2095 longer than the retry timeout, but would be bounced if a delivery was
2096 forced. I say "may" because I never totally pinned down the problem;
2097 setting up timeout/retry tests is difficult. See also the next item.
209979. The ultimate address timeout was not being applied to errors that involved
2100 a combination of host plus message (for example, a timeout on a MAIL
2101 command). When an address resolved to a number of possible hosts, and they
2102 were not all tried for each delivery (e.g. because of hosts_max_try), a
2103 message could remain on the queue longer than the retry timeout.
210580. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2106 Haardt.
210881. Fixed an obscure SMTP outgoing bug which required at least the following
2109 conditions: (a) there was another message waiting for the same server;
2110 (b) the server returned 5xx to all RCPT commands in the first message so
2111 that the message was not completed; (c) the server dropped the connection
2112 or gave a negative response to the RSET that Exim sends to abort the
2113 transaction. The observed case was a dropped connection after DATA that had
2114 been sent in pipelining mode. That is, the server had advertised PIPELINING
2115 but was not implementing it correctly. The effect of the bug was incorrect
2116 behaviour, such as trying another host, and this could lead to a crash.
2119Exim version 4.30
2122 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2123 and daemon.c were passed as pointers to ints; they should have been
2124 pointers to socklen_t variables (which are typically unsigned ints).
2126 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2127 fixed.
2129 3. Fixed a really odd bug that affected only the testing scheme; patching a
2130 certain fixed string in the binary changed the value of another string that
2131 happened to be identical to the end of the original first string.
2133 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2134 name", it returns that address as the IP address. On some operating
2135 systems (e.g. Solaris), it also passes back the IP address string as the
2136 "host name". However, on others (e.g. Linux), it passes back an empty
2137 string. Exim wasn't checking for this, and was changing the host name to an
2138 empty string, assuming it had been canonicized.
2140 5. Although rare, it is permitted to have more than one PTR record for a given
2141 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2142 all the names associated with an address, because they do in Solaris.
2143 However, it seems that they do not in Linux for data that comes from the
2144 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2145 I found this out when I moved to a new Linux workstation and tried to run
2146 the Exim test suite.
2148 To get round this problem I have changed the code so that it now does its
2149 own call to the DNS to look up PTR records when searching for a host name.
2150 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2151 addresses that are only in /etc/hosts are still found.
2153 This behaviour is, however, controlled by an option called host_lookup_
2154 order, which defaults to "bydns:byaddr". If people want to use the other
2155 order, or indeed, just use one or the other means of lookup, they can
2156 specify it in this variable.
2158 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2159 some operating systems, this comes back from gethostbyaddr() as an empty
2160 string, and this is what Exim used to test for. However, it seems that in
2161 other systems, "." is yielded. Exim now tests for this case too.
2163 7. The values of check_spool_space and check_log_space are now held internally
2164 as a number of kilobytes instead of an absolute number of bytes. If a
2165 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2166 kilobyte. This means that much larger values can be stored.
2168 8. Exim monitor: an attempt to get the action menu when not actually pointing
2169 at a message produces an empty menu entitled "No message selected". This
2170 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2171 no entries in it ("Shell widget menu has zero width and/or height"). So I
2172 have added a single, blank menu entry in this case.
2174 9. Added ${quote_local_part.
217610. MIME decoding is now applied to the contents of Subject: header lines when
2177 they are logged.
217911. Now that a reference to $sender_host_address automatically causes a reverse
2180 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2181 host lookup before query-style lookups in lists that might use this
2182 variable. This has therefore been abolished, and the "net-" prefix is no
2183 longer necessary for query-style lookups.
218512. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2186 have been a typo for LFLAGS, so it has been changed.
218813. The install script calls Exim with "-C /dev/null" in order to find the
2189 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2190 to be output. Howeve, since Exim outputs its version number before the
2191 error, it didn't break the script. It just looked ugly. I fixed this by
2192 always allowing "-C /dev/null" if the caller is root.
219414. Ignore overlarge ACL variable number when reading spool file - insurance
2195 against a later release with more variables having written the file.
219715. The standard form for an IPv6 address literal was being rejected by EHLO.
2198 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2199 this, as well as the form without the "IPv6" on the front.
220116. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2202 OS/Makefile-Darwin file.
220417. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2205 only to LDAP libraries that do not have LDAP_OPT_DEREF.
220718. After change 4.21/52, "%ld" was used to format the contents of the $inode
2208 variable. However, some OS use ints for inodes. I've added cast to long int
2209 to get rid of the compiler warning.
221119. I had forgotten to lock out "/../" in configuration file names when
2212 ALT_CONFIG_PREFIX was set.
221420. Routers used for verification do not need to specify transports. However,
2215 if such a router generated a host list, and callout was configured, Exim
2216 crashed, because it could not find a port number from the (non-existent)
2217 transport. It now assumes port 25 in this circumstance.
221921. Added the -t option to exigrep.
222122. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2222 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2223 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2224 them is set, LOOKUP_LSEARCH is forced.
222623. "exim -bV" now outputs a list of lookups that are included in the binary.
222824. Added sender and host information to the "rejected by local_scan()" log
2229 line; previously there was no indication of these.
223125. Added .include_if_exists.
223326. Change 3.952/11 added an explicit directory sync on top of a file sync for
2234 Linux. It turns out that not all file systems support this. Apparently some
2235 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2236 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2237 not supported on the file descriptor, is now ignored when Exim is trying to
2238 sync a directory. This applies only to Linux.
224027. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
224228. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2243 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2244 this.
224629. Added support for maildirsize files from supplied patch (modified a bit).
224830. The use of :fail: followed by an empty string could lead Exim to respond to
2249 sender verification failures with (e.g.):
2251 550 Verification failed for <xxx>
2252 550 Sender verify failed
2254 where the first response line was missing the '-' that indicates it is not
2255 the final line of the response.
225731. The loop for finding the name of the user that called Exim had a hardwired
2258 limit of 10; it now uses the value of finduser_retries, which is used for
2259 all other user lookups.
226132. Added $received_count variable, available in data and not_smtp ACLs, and at
2262 delivery time.
226433. Exim was neglecting to zero errno before one call of strtol() when
2265 expanding a string and expecting an integer value. On some systems this
2266 resulted in spurious "integer overflow" errors. Also, it was casting the
2267 result into an int without checking.
226934. Testing for a connection timeout using "timeout_connect" in the retry rules
2270 did not work. The code looks as if it has *never* worked, though it appears
2271 to have been documented since at least releast 1.62. I have made it work.
227335. The "timeout_DNS" error in retry rules, also documented since at least
2274 1.62, also never worked. As it isn't clear exactly what this means, and
2275 clearly it isn't a major issue, I have abolished the feature by treating it
2276 as "timeout", and writing a warning to the main and panic logs.
227836. The display of retry rules for -brt wasn't always showing the error code
2279 correctly.
228137. Added new error conditions to retry rules: timeout_A, timeout_MX,
2282 timeout_connect_A, timeout_connect_MX.
228438. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2285 to the empty sender.
228739. The daemon was not analysing the content of -oX till after it had closed
2288 stderr and disconnected from the controlling terminal. This meant that any
2289 syntax errors were only noted on the panic log, and the return code from
2290 the command was 0. By re-arranging the code a little, I've made the
2291 decoding happen first, so such errors now appear on stderr, and the return
2292 code is 1. However, the actual setting up of the sockets still happens in
2293 the disconnected process, so errors there are still only recorded on the
2294 panic log.
229640. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2297 connections (as happens on some IP stacks) was logged at start up time as
2298 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2299 it from "IPv6 and IPv4", which means that two separate sockets are being
2300 used.
230241. The debug output for gethostbyname2() or getipnodebyname() failures now
2303 says whether AF_INET or AF_INET6 was passed as an argument.
230542. Exiwhat output was messed up when time zones were included in log
2306 timestamps.
230843. Exiwhat now gives more information about the daemon's listening ports,
2309 and whether -tls-on-connect was used.
231144. The "port" option of the smtp transport is now expanded.
231345. A "message" modifier in a "warn" statement in a non-message ACL was being
2314 silently ignored. Now an error message is written to the main and panic
2315 logs.
231746. There's a new ACL modifier called "logwrite" which writes to a log file
2318 as soon as it is encountered.
232047. Added $local_user_uid and $local_user_gid at routing time.
232248. Exim crashed when trying to verify a sender address that was being
2323 rewritten to "<>".
232549. Exim was recognizing only a space character after ".include". It now also
2326 recognizes a tab character.
232850. Fixed several bugs in the Perl script that creates the exim.8 man page by
2329 extracting the relevant information from the specification. The man page no
2330 longer contains scrambled data for the -d option, and I've added a section
2331 at the front about calling Exim under different names.
233351. Added "extra_headers" argument to the "mail" command in filter files.
233552. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2336 crash.
233853. Installed eximstats 1.29.
234054. Added transport_filter_timeout as a generic transport option.
234255. Exim no longer adds an empty Bcc: header to messages that have no To: or
2343 Cc: header lines. This was required by RFC 822, but it not required by RFC
2344 2822.
234656. Exim used to add From:, Date:, and Message-Id: header lines to any
2347 incoming messages that did not have them. Now it does so only if the
2348 message originates locally, that is, if there is no associated remote host
2349 address. When Resent- header lines are present, this applies to the Resent-
2350 lines rather than the non-Resent- lines.
235257. Drop incoming SMTP connection after too many syntax or protocol errors. The
2353 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
235558. Messages for configuration errors now include the name of the main
2356 configuration file - useful now that there may be more than one file in a
2357 list (.included file names were always shown).
235959. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2360 for those rare installations that do not start the daemon as root or run it
2361 setuid root. I've cut out the call to initgroups() if the daemon is not
2362 root at that time.
236460. The Exim user and group can now be bound into the binary as text strings
2365 that are looked up at the start of Exim's processing.
236761. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
236962. Added $mailstore_basename variable.
237163. Installed patch to sieve.c from Michael Haardt.
237364. When Exim failed to open the panic log after failing to open the main log,
2374 the original message it was trying to log was written to stderr and debug
2375 output, but if they were not available (the usual case in production), it
2376 was lost. Now it is written to syslog before the two lines that record the
2377 failures to open the logs.
237965. Users' Exim filters run in subprocesses under the user's uid. It is
2380 possible for a "deliver" command or an alias in a "personal" command to
2381 provoke an address rewrite. If logging of address rewriting is configured,
2382 this fails because the process is not running as root or exim. There may be
2383 a better way of dealing with this, but for the moment (because 4.30 needs
2384 to be released), I have disabled address rewrite logging when running a
2385 filter in a non-root, non-exim process.
2388Exim version 4.24
2391 1. The buildconfig auxiliary program wasn't quoting the value set for
2392 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2393 not defined. This bug was masked in 4.22 by the effect that was fixed in
2394 change 4.23/1.
2396 2. Some messages that were rejected after a message id was allocated were
2397 shown as "incomplete" by exigrep. It no longer does this for messages that
2398 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2400 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2401 have allow_domain_literals set, the ID did not get logged in the <= line.
2402 Domain literals are now always recognized in Message-ID: header lines.
2404 4. The first argument for a ${extract expansion item is the key name or field
2405 number. Leading and trailing spaces in this item were not being ignored,
2406 causing some misleading effects.
2408 5. When deliver_drop_privilege was set, single queue runner processes started
2409 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2410 same command in the process it spins off) were not dropping privilege.
2412 6. When the daemon running as "exim" started a queue runner, it always
2413 re-executed Exim in the spun-off process. This is a waste of effort when
2414 deliver_drop_privilege is set. The new process now just calls the
2415 queue-runner function directly.
2418Exim version 4.23
2421 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2424 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2425 ignored. Though the use of -oP was forcing the writing of a pid file, it
2426 was always written to the default place.
2428 3. If the message "no IP address found for host xxxx" is generated during
2429 incoming verification, it is now followed by identification of the incoming
2430 connection (so you can more easily find what provoked it).
2432 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2434 5. Added some features to "harden" Exim a bit more against certain attacks:
2436 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2437 be put in Local/Makefile. This is like the never_users runtime option,
2438 but it cannot be overridden. The default setting is "root".
2440 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2441 prefix string with which any file named in a -C command line option
2442 must start.
2444 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2445 is retained for -C and -D only if the caller of Exim is root. Without
2446 it, the exim user may also use -C and -D and retain privilege.
2448 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2449 command line option is disabled.
2451 6. Macro names set by the -D option must start with an upper case letter, just
2452 like macro names defined in the configuration file.
2454 7. Added "dereference=" facility to LDAP.
2456 8. Two instances of the typo "uknown" in the source files are fixed.
2458 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2459 the Configure-Makefile script screwed up while processing it.
246110. Incorporated PCRE 4.4.
246311. The SMTP synchronization check was not operating right at the start of an
2464 SMTP session. For example, it could not catch a HELO sent before the client
2465 waited for the greeting. There is now a check for outstanding input at the
2466 point when the greeting is written. Because of the duplex, asynchronous
2467 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2468 way, but not yet received, when the check is performed.
247012. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2471 on TCP/IP sockets, because this apparently causes some broken clients to
2472 timeout.
247413. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2475 unchanged) from the Cygwin maintainer.
247714. The code for -bV that shows what is in the binary showed "mbx" when maildir
2478 was supported instead of testing for mbx. Effectively a typo.
248015. The spa authenticator server code was not checking that the input it
2481 received was valid base64.
248316. The debug output line for the "set" modifier in ACLs was not showing the
2484 name of the variable that was being set.
248617. Code tidy: the variable type "vtype_string" was never used. Removed it.
248818. Previously, a reference to $sender_host_name did not cause a DNS reverse
2489 lookup on its own. Something else was needed to trigger the lookup. For
2490 example, a match in host_lookup or the need for a host name in a host list.
2491 Now, if $sender_host_name is referenced and the host name has not yet been
2492 looked up, a lookup is performed. If the lookup fails, the variable remains
2493 empty, and $host_lookup_failed is set to "1".
249519. Added "eqi" as a case-independent comparison operator.
249720. The saslauthd authentication condition could segfault if neither service
2498 nor realm was specified.
250021. If an overflowing value such as "2048M" was set for message_size_limit, the
2501 error message that was logged was misleading, and incoming SMTP
2502 connections were dropped. The message is now more accurate, and temporary
2503 errors are given to SMTP connections.
250522. In some error situations (such as 21 above) Exim rejects all SMTP commands
2506 (except RSET) with a 421 error, until QUIT is received. However, it was
2507 failing to send a response to QUIT.
250923. The HELO ACL was being run before the code for helo_try_verify_hosts,
2510 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2511 ACL is now run after the helo_try_verify_hosts code.
251324. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2514 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2515 case-independent, so other case variants are also recognized). Apparently
2516 some systems use these upper case variants.
251825. If more than two messages were waiting for the same host, and a transport
2519 filter was specified for the transport, Exim sent two messages over the
2520 same TCP/IP connection, and then failed with "socket operation on non-
2521 socket" when it tried to send the third.
252326. Added Exim::debug_write and Exim::log_write for embedded Perl use.
252527. The extern definition of crypt16() in expand.c was not being excluded when
2526 the OS had its own crypt16() function.
252828. Added bounce_return_body as a new option, and bounce_return_size_limit
2529 as a preferred synonym for return_size_limit, both as an option and as an
2530 expansion variable.
253229. Added LIBS=-liconv to OS/Makefile-OSF1.
253430. Changed the default configuration ACL to relax the local part checking rule
2535 for addresses that are not in any local domains. For these addresses,
2536 slashes and pipe symbols are allowed within local parts, but the sequence
2537 /../ is explicitly forbidden.
253931. SPA server authentication was not clearing the challenge buffer before
2540 using it.
254232. log_message in a "warn" ACL statement was writing to the reject log as
2543 well as to the main log, which contradicts the documentation and doesn't
2544 seem right (because no rejection is happening). So I have stopped it.
254633. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2547 However, I am unable to do any testing of this.
254934. Fixed an infelicity in the appendfile transport. When checking directories
2550 for a mailbox, to see if any needed to be created, it was accidentally
2551 using path names with one or more superfluous leading slashes; tracing
2552 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
255435. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2555 contents are added to the log line that is written for every discarded
2556 recipient. (Previously a log_message setting was ignored.)
255836. The ${quote: operator now quotes the string if it is empty.
256037. The install script runs exim in order to find its version number. If for
2561 some reason other than non-existence or emptiness, which it checks, it
2562 could not run './exim', it was installing it with an empty version number,
2563 i.e. as "exim-". This error state is now caught, and the installation is
2564 aborted.
256638. An argument was missing from the function that creates an error message
2567 when Exim fails to connect to the socket for saslauthd authentication.
2568 This could cause Exim to crash, or give a corrupted message.
257039. Added isip, isip4, and isip6 to ${if conditions.
257240. The ACL variables $acl_xx are now saved with the message, and can be
2573 accessed later in routers, transports, and filters.
257541. The new lookup type nwildlsearch is like wildlsearch, except that the key
2576 strings in the file are not string-expanded.
257842. If a MAIL command specified a SIZE value that was too large to fit into an
2579 int variable, the check against message_size_limit failed. Such values are
2580 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2581 day this will have to be increased, but I don't think I want to be around
2582 when emails are that large.
2586Exim version 4.22
2589 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2590 iconv() is not standard in FreeBSD.
2592 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2593 IPv6 enabled. The observed symptom was a segmentation fault on return from
2594 the function os_common_find_running_interfaces() in src/os.c.
2596 3. In the check_special_case() function in daemon.c I had used "errno" as an
2597 argument name, which causes warnings on some systems. This was basically a
2598 typo, since it was named "eno" in the comments!
2600 4. The code that waits for the clock to tick (at a resolution of some fraction
2601 of a second) so as to ensure message-id uniqueness was always waiting for
2602 at least one whole tick, when it could have waited for less. [This is
2603 almost certainly not relevant at current processor speeds, where it is
2604 unlikely to ever wait at all. But we try to future-proof.]
2606 5. The function that sleeps for a time interval that includes fractions of a
2607 second contained a race. It did not block SIGALRM between setting the
2608 timer, and suspending (a couple of lines later). If the interval was short
2609 and the sigsuspend() was delayed until after it had expired, the suspension
2610 never ended. On busy systems this could lead to processes getting stuck for
2611 ever.
2613 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2614 process, before it forks any delivery processes. The open lookup caching
2615 mechanism meant that the open file or database connection was passed into
2616 the delivery process. The problem was that delivery processes always tidy
2617 up cached lookup data. This could cause a problem for the next delivery
2618 process started by the queue runner, because the external queue runner
2619 process does not know about the closure. So the next delivery process
2620 still has data in the lookup cache. In the case of a file lookup, there was
2621 no problem because closing a file descriptor in a subprocess doesn't affect
2622 the parent. However, if the lookup was caching a connection to a database,
2623 the connection was closed, and the second delivery process was likely to
2624 see errors such as "PGSQL: query failed: server closed the connection
2625 unexpectedly". The problem has been fixed by closing all cached lookups
2626 in a queue runner before running a delivery process.
2628 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2629 seem to have the "const" qualifier which it has on other OS. I've
2630 parameterised it.
2632 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2633 *of the same type* (client or server) with the same public name that an
2634 error should be diagnosed.
2636 9. When Exim looked up a host name for an IP address, but failed to find the
2637 original IP address when looking up the host name (a safety check), it
2638 output the message "<ip address> does not match any IP for NULL", which was
2639 confusing, to say the least. The bug was that the host name should have
2640 appeared instead of "NULL".
264210. Since release 3.03, if Exim is called by a uid other than root or the Exim
2643 user that is built into the binary, and the -C or -D options is used, root
2644 privilege is dropped before the configuration file is read. In addition,
2645 logging is switched to stderr instead of the normal log files. If the
2646 configuration then re-defines the Exim user, the unprivileged environment
2647 is probably not what is expected, so Exim logs a panic warning message (but
2648 proceeds).
2650 However, if deliver_drop_privilege is set, the unprivileged state may well
2651 be exactly what is intended, so the warning has been cut out in that case,
2652 and Exim is allowed to try to write to its normal log files.
2655Exim version 4.21
2658 1. smtp_return_error_details was not giving details for temporary sender
2659 or receiver verification errors.
2661 2. Diagnose a configuration error if two authenticators have the same public
2662 name.
2664 3. Exim used not to create the message log file for a message until the first
2665 delivery attempt. This could be confusing when incoming messages were held
2666 for policy or load reasons. The message log file is now created at the time
2667 the message is received, and an initial "Received" line is written to it.
2669 4. The automatically generated man page for command line options had a minor
2670 bug that caused no ill effects; however, a more serious problem was that
2671 the procedure for building the man page automatically didn't always
2672 operate. Consequently, release 4.20 contains an out-of-date version. This
2673 shouldn't happen again.
2675 5. When building Exim with embedded Perl support, the script that builds the
2676 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2677 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2679 6. The freeze_tell option was not being used for messages that were frozen on
2680 arrival, either by an ACL or by local_scan().
2682 7. Added the smtp_incomplete_transaction log selector.
2684 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2685 was accepting AUTH without a new EHLO.
2687 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2688 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2689 RFC.
269110. Logging of TCP/IP connections (when configured) now happens in the main
2692 daemon process instead of the child process, so that the TCP/IP connection
2693 count is more accurate (but it can never be perfect).
269511. The use of "drop" in a nested ACL was not being handled correctly in the
2696 outer ACL. Now, if condition failure induced by the nested "drop" causes
2697 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2698 or "require"), the connection is dropped.
270012. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2701 that yield "discard" can now be used with an "accept" or a "discard" verb,
2702 but an error is generated for any others (because I can't see a useful way
2703 to define what should happen).
270513. When an ACL is read dynamically from a file (or anywhere else), the lines
2706 are now processed in the same way as lines in the Exim configuration file.
2707 In particular, continuation lines are supported.
270914. Added the "dnslists = a.b.c!=n.n.n.n" feature.
271115. Added -ti meaning -t -i.
271316. Check for letters, digits, hyphens, and dots in the names of dnslist
2714 domains, and warn by logging if others are found.
271617. At least on BSD, alignment is not guarenteed for the array of ifreq's
2717 returned from GIFCONF when Exim is trying to find the list of interfaces on
2718 a host. The code in os.c has been modified to copy each ifreq to an aligned
2719 structure in all cases.
2721 Also, in some cases, the returned ifreq's were being copied to a 'struct
2722 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2723 means the last couple of bytes of an IPv6 address could be chopped if the
2724 ifreq contained only a normal sockaddr (14 bytes storage).
272618. Named domain lists were not supported in the hosts_treat_as_local option.
2727 An entry such as +xxxx was not recognized, and was treated as a literal
2728 domain name.
273019. Ensure that header lines added by a DATA ACL are included in the reject log
2731 if the ACL subsequently rejects the message.
273320. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2734 MD5 (which is deprecated).
273621. When testing a filter file using -bf, Exim was writing a message when it
2737 took the sender from a "From " line in the message, but it was not doing so
2738 when it took $return_path from a Return-Path: header line. It now does.
274022. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2741 with a valid header line field name (a series of printing characters
2742 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
274423. Changed "disc" in the source to "disk" to conform to the documentation and
2745 the book and for uniformity.
274724. Ignore Sendmail's -Ooption=value command line item.
274925. When execve() failed while trying to run a command in a pipe transport,
2750 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2751 could be confused with a return value of 69 from the command itself. This
2752 has been changed to 127, the value the shell returns if it is asked to run
2753 a non-existent command. The wording for the related log line suggests a
2754 non-existent command as the problem.
275626. If received_header_text expands to an empty string, do not add a Received:
2757 header line to the message. (Well, it adds a token one on the spool, but
2758 marks it "old" so that it doesn't get used or transmitted.)
276027. Installed eximstats 1.28 (addition of -nt option).
276228. There was no check for failure on the call to getsockname() in the daemon
2763 code. This can fail if there is a shortage of resources on the system, with
2764 ENOMEM, for example. A temporary error is now given on failure.
276629. Contrary to the C standard, it seems that in some environments, the
2767 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2768 program. Exim now does this explicitly; it affects the formatting of
2769 timestamps using strftime().
277130. If exiqsumm was given junk data, it threw up some uninitialized variable
2772 complaints. I've now initialized all the variables, to avoid this.
277432. Header lines added by a system filter were not being "seen" during
2775 transport-time rewrites.
277733. The info_callback() function passed to OpenSSL is set up with type void
2778 (*)(SSL *, int, int), as described somewhere. However, when calling the
2779 function (actually a macro) that sets it up, the type void(*)() is
2780 expected. I've put in a cast to prevent warnings from picky compilers.
278234. If a DNS black list lookup found a CNAME record, but there were no A
2783 records associated with the domain it pointed at, Exim crashed.
278535. If a DNS black list lookup returned more than one A record, Exim ignored
2786 all but the first. It now scans all returned addresses if a particular IP
2787 value is being sought. In this situation, the contents of the
2788 $dnslist_value variable are a list of all the addresses, separated by a
2789 comma and a space.
279136. Tightened up the rules for host name lookups using reverse DNS. Exim used
2792 to accept a host name and all its aliases if the forward lookup for any of
2793 them yielded the IP address of the incoming connection. Now it accepts only
2794 those names whose forward lookup yields the correct IP address. Any other
2795 names are discarded. This closes a loophole whereby a rogue DNS
2796 administrator could create reverse DNS records to break through a
2797 wildcarded host restriction in an ACL.
279937. If a user filter or a system filter that ran in a subprocess used any of
2800 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2801 the wrong values were passed to the pipe command ($thisaddress had the
2802 value of $0, $0 had the value of $1, etc). This bug was introduced by
2803 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
280538. Improved the line breaking for long SMTP error messages from ACLs.
2806 Previously, if there was no break point between 40 and 75 characters, Exim
2807 left the rest of the message alone. Two changes have been made: (a) I've
2808 reduced the minimum length to 35 characters; (b) if it can't find a break
2809 point between 35 and 75 characters, it looks ahead and uses the first one
2810 that it finds. This may give the occasional overlong line, but at least the
2811 remaining text gets split now.
281339. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2814 file descriptors might be low, and that setting 1000 would always raise it.
2815 It turns out that in some environments, the limit is already over 1000 and
2816 that lowering it causes trouble. So now Exim takes care not to decrease it.
281840. When delivering a message, the value of $return_path is set to $sender_
2819 address at the start of routing (routers may change the value). By an
2820 oversight, this default was not being set up when an address was tested by
2821 -bt or -bv, which affected the outcome if any router or filter referred to
2822 $return_path.
282441. The idea of the "warn" ACL verb is that it adds a header or writes to the
2825 log only when "message" or "log_message" are set. However, if one of the
2826 conditions was an address verification, or a call to a nested ACL, the
2827 messages generated by the underlying test were being passed through. This
2828 no longer happens. The underlying message is available in $acl_verify_
2829 message for both "message" and "log_message" expansions, so it can be
2830 passed through if needed.
283242. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2833 new expansion $bh_ to give the encoded byte string without charset
2834 translation. Translation happens only if iconv() is available; HAVE_ICONV
2835 indicates this at build time. HEADERS_CHARSET gives the charset to
2836 translate to; headers_charset can change it in the configuration, and
2837 "headers charset" can change it in an individual filter file.
283943. Now that we have a default RFC 2047 charset (see above), the code in Exim
2840 that creates RFC 2047 encoded "words" labels them as that charset instead
2841 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2842 expansion operator; (ii) when Exim creates a From: line for a local
2843 message; (iii) when a header line is rewritten to include a "phrase" part.
284544. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2846 buggy, causing it to skip the first lines of messages whose message ID
2847 ended in 'D'. This would not have bitten before Exim release 4.14, because
2848 message IDs were unlikely to end in 'D' before then. The effect was to have
2849 incorrect size information for certain domains.
285145. #include "config.h" was missing at the start of the crypt16.c module. This
2852 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2853 noticed.
285546. If there was a timeout during a "random" callout check, Exim treated it as
2856 a failure of the random address, and carried on sending RSET and the real
2857 address. If the delay was just some slowness somewhere, the response to the
2858 original RCPT would be taken as a response to RSET and so on, causing
2859 mayhem of various kinds.
286147. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2862 when I implemented it. It didn't allow for the fact that some option values
2863 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2864 the right test for positive values.
286648. Domain names in DNS records are case-independent. Exim always looks them up
2867 in lower case. Some resolvers return domain names in exactly the case they
2868 appear in the zone file, that is, they may contain uppercase letters. Not
2869 all resolvers do this - some return always lower case. Exim was treating a
2870 change of case by a resolver as a change of domain, similar to a widening
2871 of a domain abbreviation. This triggered its re-routing code and so it was
2872 trying to route what was effectively the same domain again. This normally
2873 caused routing to fail (because the router wouldn't handle the domain
2874 twice). Now Exim checks for this case specially, and just changes the
2875 casing of the domain that it ultimately uses when it transmits the message
2876 envelope.
287849. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
2879 module.
288150. If a filter generated a file delivery with a non-absolute name (possible if
2882 no home directory exists for the router), the forbid_file option was not
2883 forbidding it.
288551. Added '&' feature to dnslists, to provide bit mask matching in addition to
2886 the existing equality matching.
288852. Exim was using ints instead of ino_t variables in some places where it was
2889 dealing with inode numbers.
289153. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
2892 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
2893 called TMPDIR, and if it finds it is different, it changes its value.
289554. The smtp_printf() function is now made available to local_scan() so
2896 additional output lines can be written before returning. There is also an
2897 smtp_fflush() function to enable the detection of a dropped connection.
2898 The variables smtp_input and smtp_batched_input are exported to
2899 local_scan().
290155. Changed the default runtime configuration: the message "Unknown user"
2902 has been removed from the ACL, and instead placed on the localuser router,
2903 using the cannot_route_message feature. This means that any verification
2904 failures that generate their own messages won't get overridden. Similarly,
2905 the "Unrouteable address" message that was in the ACL for unverifiable
2906 relay addresses has also been removed.
290856. Added hosts_avoid_esmtp to the smtp transport.
291057. The exicyclog script was not checking for the esoteric option
2911 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
2912 will work only if exicyclog is run under the appropriate euid.
291458. Following a discussion on the list, the rules by which Exim recognises line
2915 endings on incoming messages have been changed. The -dropcr and drop_cr
2916 options are now no-ops, retained only for backwards compatibility. The
2917 following line terminators are recognized: LF CRLF CR. However, special
2918 processing applies to CR:
2920 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
2921 nor a local message in the state where . is a terminator.
2923 (ii) If a bare CR is encountered in a header line, an extra space is added
2924 after the line terminator so as not to end the header. The reasoning
2925 behind this is that bare CRs in header lines are most likely either
2926 to be mistakes, or people trying to play silly games.
292859. The size of a message, as listed by "-bp" or in the Exim monitor window,
2929 was being incorrectly given as 18 bytes larger than it should have been.
2930 This is a VOB (very old bug).
293260. This may never have affected anything current, but just in case it has:
2933 When the local host is found other than at the start of a list of hosts,
2934 the local host, those with the same MX, and any that follow, are discarded.
2935 When the list in question was part of a longer list of hosts, the following
2936 hosts (not currently being processed) were also being discarded. This no
2937 longer happens. I'm not sure if this situation could ever has previously
2938 arisen.
294061. Added the "/MX" feature to lists of hosts in the manualroute and query
2941 program routers.
294362. Whenever Exim generates a new message, it now adds an Auto-Submitted:
2944 header. This is something that is recommended in a new Internet Draft, and
2945 is something that is documented as being done by Sendmail. There are two
2946 possible values. For messages generated by the autoreply transport, Exim
2947 adds:
2949 Auto-Submitted: auto-replied
2951 whereas for all other generated messages (e.g. bounces) it adds
2953 Auto-Submitted: auto-generated
295563. The "personal" condition in filters now includes a test for the
2956 Auto-Submitted: header. If it contains the string "auto-" the message it
2957 not considered personal.
295964. Added rcpt_include_affixes as a generic transport option.
296165. Added queue_only_override (default true).
296366. Added the syslog_duplication option.
296567. If what should have been the first header line of a message consisted of
2966 a space followed by a colon, Exim was mis-interpreting it as a header line.
2967 It isn't of course - it is syntactically invalid and should therefore be
2968 treated as the start of the message body. The misbehaviour could have
2969 caused a number of strange effects, including loss of data in subsequent
2970 header lines, and spool format errors.
297268. Formerly, the AUTH parameter on a MAIL command was trusted only if the
2973 client host had authenticated. This control can now be exercised by an ACL
2974 for more flexibility.
297669. By default, callouts do not happen when testing with -bh. There is now a
2977 variant, -bhc, which does actually run the callout code, including
2978 consulting and updating the callout cache.
298070. Added support for saslauthd authentication, courtesy of Alexander
2981 Sabourenkov.
298371. If statvfs() failed on the spool or log directories while checking their
2984 size for availability, Exim confusingly gave the error "space shortage".
2985 Furthermore, in debugging mode it crashed with a floating point exception.
2986 These checks are done if check_{spool,log}_{space,inodes} are set, and when
2987 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
2988 serious problem, Exim now writes to the main and panic logs when this
2989 happens, with details of the failure. It then refuses to accept the
2990 incoming message, giving the message "spool directory problem" or "log
2991 directory problem" with a 421 code for SMTP messages.
299372. When Exim is about to re-exec itself, it ensures that the file descriptors
2994 0, 1, and 2 exist, because some OS complain for execs without them (see
2995 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
2996 descriptors. However, the code omitted to check that the open succeeded,
2997 causing mysterious errors if for some reason the permissions on /dev/null
2998 got screwed. Now Exim writes a message to the main and panic logs, and
2999 bombs out if it can't open /dev/null.
300173. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3002 interact so that it is all more flexible. It is supposed to remain
3003 backwards compatible. Also added extra_local_interfaces.
300574. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3006 to bomb out with an assertion failure - to the client this appears as a
3007 connection drop. This problem occurs in the part of the code that was taken
3008 from the Samba project. Fortunately, the assertion is in a very simple
3009 function, so I have fixed this by reproducing the function inline in the
3010 one place where it is called, and arranging for authentication to fail
3011 instead of killing the process with assert().
301375. The SPA client code was not working when the server requested OEM rather
3014 than Unicode encoding.
301676. Added code to make require_files with a specific uid setting more usable in
3017 the case where statting the file as root fails - usually a non-root-mounted
3018 NFS file system. When this happens and the failure is EACCES, Exim now
3019 forks a subprocess and does the per-uid checking as the relevant uid.
302177. Added process_log_path.
302378. If log_file_path was not explicitly set, a setting of check_log_space or
3024 check_log_inodes was ignored.
302679. If a space check for the spool or log partitions fails, the incident is now
3027 logged. Of course, in the latter case the data may get lost...
302980. Added the %p formatting code to string_format() so that it can be used to
3030 print addresses in debug_print(). Adjusted all the address printing in the
3031 debugging in store.c to use %p rather than %d.
303381. There was a concern that a line of code in smtp_in.c could overflow a
3034 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3035 initially expressed, the concern was not well-founded, because trailing
3036 spaces are removed early. However, if the trailing spaces were followed by
3037 a NULL, they did not get removed, so the overflow was possible. Two fixes
3038 were applied:
3040 (a) I re-wrote the offending code in a cleaner fashion.
3041 (b) If an incoming SMTP command contains a NULL character, it is rejected
3042 as invalid.
304482. When Exim changes uid/gid to the Exim user at daemon start time, it now
3045 runs initgroups(), so that if the Exim user is in any additional groups,
3046 they will be used during message reception.
3049Exim version 4.20
3052The change log for 4.20 and earlier releases has been archived.