Follow links when building the tarballs for config.samples (these
[exim.git] / doc / doc-txt / ChangeLog
35af9f61 1$Cambridge: exim/doc/doc-txt/ChangeLog,v 1.5 2004/10/14 14:52:45 ph10 Exp $
3Change log file for Exim from version 4.21
7Exim version 4.44
10 1. Minor wording change to the doc/README.SIEVE file.
139059f6 12 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5fe762f6 13 computation of the current number of files was incorrect.
495ae4b0 14
15 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
16 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
17 place.
19 4. Give more explanation in the error message when the command for a transport
20 filter fails to execute.
23Exim version 4.43
26 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
27 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
28 Now it returns an integer. A number of calls were still expecting a T/F
29 return. Fortuitously, in all cases, the tests worked in OK situations,
30 which is the norm. However, things would have gone wrong on any write
31 failures on the smtp file descriptor. This function is used when sending
32 messages over SMTP and also when doing verify callouts.
34 2. When Exim is called to do synchronous delivery of a locally submitted
35 message (the -odf or -odi options), it no longer closes stderr before doing
36 the delivery.
38 3. Implemented the mua_wrapper option.
40 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
42 5. Implemented the functions header_remove(), header_testname(),
43 header_add_at_position(), and receive_remove_recipient(), and exported them
44 to local_scan().
46 6. If an ACL "warn" statement specified the addition of headers, Exim already
47 inserted X-ACL-Warn: at the start if there was no header name. However, it
48 was not making this test for the second and subsequent header lines if
49 there were newlines in the string. This meant that an invalid header could
50 be inserted if Exim was badly configured.
52 7. Allow an ACL "warn" statement to add header lines at the start or after all
53 the Received: headers, as well as at the end.
55 8. Added the rcpt_4xx retry error code.
57 9. Added postmaster_mailfrom=xxx to callout verification option.
5910. Added mailfrom=xxxx to the callout verification option, for verify=
60 header_sender only.
6211. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
63 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6512. Inserted some casts to stop certain compilers warning when using pointer
66 differences as field lengths or precisions in printf-type calls (mostly
67 affecting debugging statements).
6913. Added optional readline() support for -be (dynamically loaded).
7114. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
72 same clock tick as a message's arrival, so that its received time was the
73 same as the "first fail" time on the retry record, and that message
74 remained on the queue past the ultimate address timeout, every queue runner
75 would try a delivery (because it was past the ultimate address timeout) but
76 after another failure, the ultimate address timeout, which should have then
77 bounced the address, did not kick in. This was a "< instead of <=" error;
78 in most cases the first failure would have been in the next clock tick
79 after the received time, and all would be well.
8115. The special items beginning with @ in domain lists (e.g. @mx_any) were not
82 being recognized when the domain list was tested by the match_domain
83 condition in an expansion string.
8516. Added the ${str2b64: operator.
8717. Exim was always calling setrlimit() to set a large limit for the number of
88 processes, without checking whether the existing limit was already
89 adequate. (It did check for the limit on file descriptors.) Furthermore,
90 errors from getrlimit() and setrlimit() were being ignored. Now they are
91 logged to the main and panic logs, but Exim does carry on, to try to do its
92 job under whatever limits there are.
9418. Imported PCRE 5.0.
9619. Trivial typo in log message " temporarily refused connection" (the leading
97 space).
9920. If the log selector return_path_on_delivery was set and an address was
100 redirected to /dev/null, the delivery process crashed because it assumed
101 that a return path would always be set for a "successful" delivery. In this
102 case, the whole delivery is bypassed as an optimization, and therefore no
103 return path is set.
10521. Internal re-arrangement: the function for sending a challenge and reading
106 a response while authentication was assuming a zero-terminated challenge
107 string. It's now changed to take a pointer and a length, to allow for
108 binary data in such strings.
11022. Added the cyrus_sasl authenticator (code supplied by MBM).
11223. Exim was not respecting finduser_retries when seeking the login of the
113 uid under which it was called; it was always trying 10 times. (The default
114 setting of finduser_retries is zero.) Also, it was sleeping after the final
115 failure, which is pointless.
11724. Implemented tls_on_connect_ports.
11925. Implemented acl_smtp_predata.
12126. If the domain in control=submission is set empty, Exim assumes that the
122 authenticated id is a complete email address when it generates From: or
123 Sender: header lines.
12527. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
126 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
127 chown and chgrp in /bin and hostname in /usr/bin.
12928. Exim was keeping the "process log" file open after each use, just as it
130 does for the main log. This opens the possibility of it remaining open for
131 long periods when the USR1 signal hits a daemon. Occasional processlog
132 errors were reported, that could have been caused by this. Anyway, it seems
133 much more sensible not to leave this file open at all, so that is what now
134 happens.
13629. The long-running daemon process does not normally write to the log once it
137 has entered its main loop, and it closes the log before doing so. This is
138 so that log files can straightforwardly be renamed and moved. However,
139 there are a couple of unusual error situations where the daemon does write
140 log entries, and I had neglected to close the log afterwards.
14230. The text of an SMTP error response that was received during a remote
143 delivery was being truncated at 512 bytes. This is too short for some of
144 the long messages that one sometimes sees. I've increased the limit to
145 1024.
14731. It is now possible to make retry rules that apply only when a message has a
148 specific sender, in particular, an empty sender.
15032. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
151 it possible to be selective about when SMTP synchronization is enforced.
15333. Added "control = caseful_local_part" and "control = "caselower_local_part".
15532. Implemented hosts_connection_nolog.
15733. Added an ACL for QUIT.
15934. Setting "delay_warning=" to disable warnings was not working; it gave a
160 syntax error.
16235. Added mailbox_size and mailbox_filecount to appendfile.
16436. Added control = no_multiline_responses to ACLs.
16637. There was a bug in the logic of the code that waits for the clock to tick
167 in the case where the clock went backwards by a substantial amount such
168 that the microsecond fraction of "now" was more than the microsecond
169 fraction of "then" (but the whole seconds number was less).
17138. Added support for the libradius Radius client library this is found on
172 FreeBSD (previously only the radiusclient library was supported).
175Exim version 4.42
178 1. When certain lookups returned multiple values in the form name=value, the
179 quoting of the values was not always being done properly. Specifically:
180 (a) If the value started with a double quote, but contained no whitespace,
181 it was not quoted.
182 (b) If the value contained whitespace other than a space character (i.e.
183 tabs or newlines or carriage returns) it was not quoted.
184 This fix has been applied to the mysql and pgsql lookups by writing a
185 separate quoting function and calling it from the lookup code. The fix
186 should probably also be applied to nisplus, ibase and oracle lookups, but
187 since I cannot test any of those, I have not disturbed their existing code.
189 2. A hit in the callout cache for a specific address caused a log line with no
190 reason for rejecting RCPT. Now it says "Previous (cached) callout
191 verification failure".
193 3. There was an off-by-one bug in the queryprogram router. An over-long
194 return line was truncated at 256 instead of 255 characters, thereby
195 overflowing its buffer with the terminating zero. As well as fixing this, I
196 have increased the buffer size to 1024 (and made a note to document this).
198 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
199 when Exim is waiting for an SMTP response from a remote server, Exim
200 restarts its select() call on the socket, thereby resetting its timeout.
201 This is not a problem when such interrupts are rare. Somebody set up a cron
202 job to run exiwhat every 2 minutes, which is less than the normal select()
203 timeout (5 or 10 minutes). This meant that the select() timeout never
204 kicked in because it was always reset. I have fixed this by comparing the
205 time when an interrupt arrives with the time at the start of the first call
206 to select(). If more time than the timeout has elapsed, the interrupt is
207 treated as a timeout.
209 5. Some internal re-factoring in preparation for the addition of Sieve
210 extensions (by MH). In particular, the "personal" test is moved to a
211 separate function, and given an option for scanning Cc: and Bcc: (which is
212 not set for Exim filters).
214 6. When Exim created an email address using the login of the caller as the
215 local part (e.g. when creating a From: or Sender: header line), it was not
216 quoting the local part when it contained special characters such as @.
218 7. Installed new OpenBSD configuration files.
220 8. Reworded some messages for syntax errors in "and" and "or" conditions to
221 try to make them clearer.
223 9. Callout options, other than the timeout value, were being ignored when
224 verifying sender addresses in header lines. For example, when using
226 verify = header_sender/callout=no_cache
228 the cache was (incorrectly) being used.
23010. Added a missing instance of ${EXE} to the exim_install script; this affects
231 only the Cygwin environment.
23311. When return_path_on_delivery was set as a log selector, if different remote
234 addresses in the same message used different return paths and parallel
235 remote delivery occurred, the wrong values would sometimes be logged.
236 (Whenever a remote delivery process finished, the return path value from
237 the most recently started remote delivery process was logged.)
23912. RFC 3848 specifies standard names for the "with" phrase in Received: header
240 lines when AUTH and/or TLS are in use. This is the "received protocol"
241 field. Exim used to use "asmtp" for authenticated SMTP, without any
242 indication (in the protocol name) for TLS use. Now it follows the RFC and
243 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
244 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
245 names appear in log lines as well as in Received: header lines.
24713. Installed MH's patches for Sieve to add the "copy" and "vacation"
248 extensions, and comparison tests, and to fix some bugs.
25014. Changes to the "personal" filter test:
252 (1) The test was buggy in that it was just doing the equivalent of
253 "contains" tests on header lines. For example, if a user's address was
254 anne@some.where, the "personal" test would incorrectly be true for
256 To: susanne@some.where
258 This test is now done by extracting each address from the header in turn,
259 and checking the entire address. Other tests that are part of "personal"
260 are now done using regular expressions (for example, to check local parts
261 of addresses in From: header lines).
263 (2) The list of non-personal local parts in From: addresses has been
264 extended to include "listserv", "majordomo", "*-request", and "owner-*",
265 taken from the Sieve specification recommendations.
267 (3) If the message contains any header line starting with "List-" it is
268 treated as non-personal.
270 (4) The test for "circular" in the Subject: header line has been removed
271 because it now seems ill-conceived.
27315. Minor typos in src/EDITME comments corrected.
27516. Installed latest exipick from John Jetmore.
27717. If headers_add on a router specified a text string that was too long for
278 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
279 of string_sprintf() is now avoided.
28118. $message_body_size was not set (it was always zero) when running the DATA
282 ACL and the local_scan() function.
28419. For the "mail" command in an Exim filter, no default was being set for
285 the once_repeat time, causing a random time value to be used if "once" was
286 specified. (If the value happened to be <= 0, no repeat happened.) The
287 default is now 0s, meaning "never repeat". The "vacation" command was OK
288 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
289 (I found it when inspecting the code).
29120. There is now an overall timeout for performing a callout verification. It
292 defaults to 4 times the callout timeout, which applies to individual SMTP
293 commands during the callout. The overall timeout applies when there is more
294 than one host that can be tried. The timeout is checked before trying the
295 next host. This prevents very long delays if there are a large number of
296 hosts and all are timing out (e.g. when the network connections are timing
297 out). The value of the overall timeout can be changed by specifying an
298 additional sub-option for "callout", called "maxwait". For example:
300 verify = sender/callout=5s,maxwait=20s
30221. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
303 to the end before writing, but this should make it even safer).
30522. Exim was forgetting that it had advertised PIPELINING for the second and
306 subsequent messages on an SMTP connection. It was also not resetting its
307 memory on STARTTLS and an internal HELO.
30923. When Exim logs an SMTP synchronization error within a session, it now
310 records whether PIPELINING has been advertised or not.
31224. Added 3 instances of "(long int)" casts to time_t variables that were being
313 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
314 rather than long int.
31625. Installed the latest Cygwin configuration files from the Cygwin maintainer.
31826. Added the never_mail option to autoreply.
321Exim version 4.41
324 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
325 crash if the getsockname() call failed; this can happen if a connection is
326 closed very soon after it is established. The problem was simply in the
327 order in which certain operations were done, causing Exim to try to write
328 to the SMTP stream before it had set up the file descriptor. The bug has
329 been fixed by making things happen in the correct order.
332Exim version 4.40
335 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
336 before the connection was closed, thus losing the rejection response.
338 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
339 some early Solaris releases, but causes trouble in current releases where
340 socklen_t is defined.
342 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
343 always exist.
345 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
346 configured.
348 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
349 that releasing the top of it at the end releases what was used for sub-
350 expansions (unless the block got too big). However, discard this block if
351 the first thing is a variable or header, so that we can use its block when
352 it is dynamic (useful for very large $message_headers, for example).
354 6. Lookups now cache *every* query, not just the most recent. A new, separate
355 store pool is used for this. It can be recovered when all lookup caches are
356 flushed. Lookups now release memory at the end of their result strings.
357 This has involved some general refactoring of the lookup sources.
359 7. Some code has been added to the store_xxx() functions to reduce the amount
360 of flapping under certain conditions.
362 8. log_incoming_interface used to affect only the <= reception log lines. Now
363 it causes the local interface and port to be added to several more SMTP log
364 lines, for example "SMTP connection from", and rejection lines.
366 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
36810. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
37011. If $host_data was set by virtue of a hosts lookup in an ACL, its value
371 could be overwritten at the end of the current message (or the start of a
372 new message if it was set in a HELO ACL). The value is now preserved for
373 the duration of the SMTP connection.
37512. If a transport had a headers_rewrite setting, and a matching header line
376 contained an unqualified address, that address was qualified, even if it
377 did not match any rewriting rules. The underlying bug was that the values
378 of the flags that permit the existence of unqualified sender and recipient
379 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
380 non-local messages, and by -bnq for local messages) were not being
381 preserved with the message after it was received.
38313. When Exim was logging an SMTP synchronization error, it could sometimes log
384 "next input=" as part of the text comprising the host identity instead of
385 the correct text. The code was using the same buffer for two different
386 strings. However, depending on which order the printing function evaluated
387 its arguments, the bug did not always show up. Under Linux, for example, my
388 test suite worked just fine.
39014. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
391 doesn't work with some older versions of Perl. It has been changed to "my",
392 which in any case is probably the better facility to use.
39415. A really picky compiler found some instances of statements for creating
395 error messages that either had too many or two few arguments for the format
396 string.
39816. The size of the buffer for calls to the DNS resolver has been increased
399 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
400 for addresses that have a lot of PTR records. This alleviates a problem; it
401 does not fully solve it.
40317. A dnsdb lookup for PTR records that receives more data than will fit in the
404 buffer now truncates the list and logs the incident, which is the same
405 action as happens when Exim is looking up a host name and its aliases.
406 Previously in this situation something unpredictable would happen;
407 sometimes it was "internal error: store_reset failed".
40918. If a server dropped the connection unexpectedly when an Exim client was
410 using GnuTLS and trying to read a response, the client delivery process
411 crashed while trying to generate an error log message.
41319. If a "warn" verb in an ACL added multiple headers to a message in a single
414 string, for example:
416 warn message = H1: something\nH2: something
418 the text was added as a single header line from Exim's point of view
419 though it ended up OK in the delivered message. However, searching for the
420 second and subsequent header lines using $h_h2: did not work. This has been
421 fixed. Similarly, if a system filter added multiple headers in this way,
422 the routers could not see them.
42420. Expanded the error message when iplsearch is called with an invalid key to
425 suggest using net-iplsearch in a host list.
42721. When running tests using -bh, any delays imposed by "delay" modifiers in
428 ACLs are no longer actually imposed (and a message to that effect is
429 output).
43122. If a "gecos" field in a passwd entry contained escaped characters, in
432 particular, if it contained a \" sequence, Exim got it wrong when building
433 a From: or a Sender: header from that name. A second bug also caused
434 incorrect handling when an unquoted " was present following a character
435 that needed quoting.
43723. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
438 was not being matched caselessly.
44024. Arranged for all hyphens in the exim.8 source to be escaped with
441 backslashes.
44325. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
444 properly. Recipient callout cache records were still being keyed to include
445 the sender, even when use_sender was set false. This led to far more
446 callouts that were necessary. The sender is no longer included in the key
447 when use_sender is false.
44926. Added "control = submission" modifier to ACLs.
45127. Added the ${base62d: operator to decode base 62 numbers.
45328. dnsdb lookups can now access SRV records.
45529. CONFIGURE_OWNER can be set at build time to define an alternative owner for
456 the configuration file.
45830. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
459 (-v) mode. This makes the output for a verbose queue run more intelligible.
46131. Added a use_postmaster feature to recipient callouts.
46332. Added the $body_zerocount variable, containing the number of binary zero
464 bytes in the message body.
46633. The time of last modification of the "new" subdirectory is now used as the
467 "mailbox time last read" when there is a quota error for a maildir
468 delivery.
47034. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
47235. Added +ignore_unknown as a special item in host lists.
47436. Code for decoding IPv6 addresses in host lists is now included, even if
475 IPv6 support is not being compiled. This fixes a bug in which an IPv6
476 address was recognized as an IP address, but was then not correctly decoded
477 into binary, causing unexpected and incorrect effects when compared with
478 another IP address.
481Exim version 4.34
484 1. Very minor rewording of debugging text in manualroute to say "list of
485 hosts" instead of "hostlist".
487 2. If verify=header_syntax was set, and a header line with an unqualified
488 address (no domain) and a large number of spaces between the end of the
489 name and the colon was received, the reception process suffered a buffer
490 overflow, and (when I tested it) crashed. This was caused by some obsolete
491 code that should have been removed. The fix is to remove it!
493 3. When running in the test harness, delay a bit after writing a bounce
494 message to get a bit more predictability in the log output.
496 4. Added a call to search_tidyup() just before forking a reception process. In
497 theory, someone could use a lookup in the expansion of smtp_accept_max_
498 per_host which, without the tidyup, could leave open a database connection.
500 5. Added the variables $recipient_data and $sender_data which get set from a
501 lookup success in an ACL "recipients" or "senders" condition, or a router
502 "senders" option, similar to $domain_data and $local_part_data.
504 6. Moved the writing of debug_print from before to after the "senders" test
505 for routers.
507 7. Change 4.31/66 (moving the time when the Received: is generated) caused
508 problems for message scanning, either using a data ACL, or using
509 local_scan() because the Received: header was not generated till after they
510 were called (in order to set the time as the time of reception completion).
511 I have revised the way this works. The header is now generated after the
512 body is received, but before the ACL or local_scan() are called. After they
513 are run, the timestamp in the header is updated.
516Exim version 4.33
519 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
520 before starting a queue runner without re-exec. This happened only when
521 deliver_drop_privilege was set or when the Exim user was set to root. The
522 effect of the bug was that timeouts during subsequent deliveries caused
523 crashes instead of being properly handled. The handler is now left at its
524 default (and expected) setting.
526 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
527 message, again when deliver_drop_privilege is set or Exim is run as root.
528 The bug described in (1) was not present in this case, but the tidying up
529 of the other signals was missing. I have made the two cases consistent.
531 3. The ignore_target_hosts setting on a manualroute router was being ignored
532 for hosts that were looked up using the /MX notation.
534 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
535 in domain lists.
537 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
538 operated on the sender address. After changing the $sender_address to <>
539 for the sender address verify, Exim was re-instated it as the original
540 (before rewriting) address, but remembering that it had rewritten it, so it
541 wasn't rewriting it again. This bug also had the effect of breaking the
542 sender address verification caching when the sender address was rewritten.
544 6. The ignore_target_hosts option was being ignored by the ipliteral router.
545 This has been changed so that if the ip literal address matches
546 ignore_target_hosts, the router declines.
548 7. Added expansion conditions match_domain, match_address, and match_local_
549 part (NOT match_host).
551 8. The placeholder for the Received: header didn't have a length field set.
553 9. Added code to Exim itself and to exim_lock to test for a specific race
554 condition that could lead to file corruption when using MBX delivery. The
555 issue is with the lockfile that is created in /tmp. If this file is removed
556 after a process has opened it but before that process has acquired a lock,
557 there is the potential for a second process to recreate the file and also
558 acquire a lock. This could lead to two Exim processes writing to the file
559 at the same time. The added code performs the same test as UW imapd; it
560 checks after acquiring the lock that its file descriptor still refers to
561 the same named file.
56310. The buffer for building added header lines was of fixed size, 8192 bytes.
564 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
565 when Exim is built.
56711. Added the smtp_active_hostname option. If used, this will typically be made
568 to depend on the incoming interface address. Because $interface_address is
569 not set up until the daemon has forked a reception process, error responses
570 that can happen earlier (such as "too many connections") no longer contain
571 a host name.
57312. If an expansion in a condition on a "warn" statement fails because a lookup
574 defers, the "warn" statement is abandoned, and the next ACL statement is
575 processed. Previously this caused the whole ACL to be aborted.
57713. Added the iplsearch lookup type.
57914. Added ident_timeout as a log selector.
58115. Added tls_certificate_verified as a log selector.
58316. Added a global option tls_require_ciphers (compare the smtp transport
584 option of the same name). This controls incoming TLS connections.
58617. I finally figured out how to make tls_require_ciphers do a similar thing
587 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
588 before starting the TLS session.
59018. Tabs are now shown as \t in -bP output.
59219. If the log selector return_path_on_delivery was set, Exim crashed when
593 bouncing a message because it had too many Received: header lines.
59520. If two routers both had headers_remove settings, and the first one included
596 a superfluous trailing colon, the final name in the first list and the
597 first name in the second list were incorrectly joined into one item (with a
598 colon in the middle).
601Exim version 4.32
604 1. Added -C and -D options to the exinext utility, mainly to make it easier
605 to include in the automated testing, but these could be helpful when
606 multiple configurations are in use.
608 2. The exinext utility was not formatting the output nicely when there was
609 an alternate port involved in the retry record key, nor when there was a
610 message id as well (for retries that were specific to a specific message
611 and a specific host). It was also confused by IPv6 addresses, because of
612 the additional colons they contain. I have fixed the IPv4 problem, and
613 patched it up to do a reasonable job for IPv6.
615 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
616 delivery, the log line now contains "pipelined" if PIPELINING was used.
618 4. An SMTP transport process used to panic and die if the bind() call to set
619 an explicit outgoing interface failed. This has been changed; it is now
620 treated in the same way as a connect() failure.
622 5. A reference to $sender_host_name in the part of a conditional expansion
623 that was being skipped was still causing a DNS lookup. This no longer
624 occurs.
626 6. The def: expansion condition was not recognizing references to header lines
627 that used bh_ and bheader_.
629 7. Added the _cache feature to named lists.
631 8. The code for checking quota_filecount in the appendfile transport was
632 allowing one more file than it should have been.
634 9. For compatibility with Sendmail, the command line option
636 -prval:sval
638 is equivalent to
640 -oMr rval -oMs sval
642 and sets the incoming protocol and host name (for trusted callers). The
643 host name and its colon can be omitted when only the protocol is to be set.
644 Note the Exim already has two private options, -pd and -ps, that refer to
645 embedded Perl. It is therefore impossible to set a protocol value of "d" or
646 "s", but I don't think that's a major issue.
64810. A number of refactoring changes to the code, none of which should affect
649 Exim's behaviour:
651 (a) The number of logging options was getting close to filling up the
652 32-bit word that was used as a bit map. I have split them into two classes:
653 those that are passed in the argument to log_write(), and those that are
654 only ever tested independently outside of that function. These are now in
655 separate 32-bit words, so there is plenty of room for expansion again.
656 There is no change in the user interface or the logging behaviour.
658 (b) When building, for example, log lines, the code previously used a
659 macro that called string_cat() twice, in order to add two strings. This is
660 not really sufficiently general. Furthermore, there was one instance where
661 it was actually wrong because one of the argument was used twice, and in
662 one call a function was used. (As it happened, calling the function twice
663 did not affect the overall behaviour.) The macro has been replaced by a
664 function that can join an arbitrary number of extra strings onto a growing
665 string.
667 (c) The code for expansion conditions now uses a table and a binary chop
668 instead of a serial search (which was left over from when there were very
669 few conditions). Also, it now recognizes conditions like "pam" even when
670 the relevant support is not compiled in: a suitably worded error message is
671 given if an attempt is made to use such a condition.
67311. Added ${time_interval:xxxxx}.
67512. A bug was causing one of the ddress fields not to be passed back correctly
676 from remote delivery subprocesses. The field in question was not being
677 subsequently used, so this caused to problems in practice.
67913. Added new log selectors queue_time and deliver_time.
68114. Might have fixed a bug in maildirsizefile handling that threw up
682 "unexpected character" debug warnings, and recalculated the data
683 unnecessarily. In any case, I expanded the warning message to give more
684 information.
68615. Added the message "Restricted characters in address" to the statements in
687 the default ACL that block characters like @ and % in local parts.
68916. Change 71 for release 4.31 proved to be much less benign that I imagined.
690 Three changes have been made:
692 (a) There was a serious bug; a negative response to MAIL caused the whole
693 recipient domain to be cached as invalid, thereby blocking all messages
694 to all local parts at the same domain, from all senders. This bug has
695 been fixed. The domain is no longer cached after a negative response to
696 MAIL if the sender used is not empty.
698 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
699 been restored.
701 (c) A new callout option, "use_sender" has been added for people who want
702 the modified behaviour.
705Exim version 4.31
708 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
709 Larry Rosenman.
711 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
712 indeed breaks things for older releases.
714 3. Added additional logging to the case where there is a problem reading data
715 from a filter that is running in a subprocess using a pipe, in order to
716 try to track down a specific problem.
718 4. Testing facility fudge: when running in the test harness and attempting
719 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
720 getting "No route to host". Convert this to a timeout.
722 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
723 warning.
725 6. Some OS don't have socklen_t but use size_t instead. This affects the
726 fifth argument of getsockopt() amongst other things. This is now
727 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
728 can be set for individual OS. I have set it for SunOS5, OSF1, and
729 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
730 some earlier ones do not.
732 7. Change 4.30/15 was not doing the test caselessly.
734 8. The standard form for an IPv6 address literal was being rejected by address
735 parsing in, for example, MAIL and RCPT commands. An example of this kind of
736 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
737 this, as well as the form without the "IPv6" on the front (but only when
738 address literals are enabled, of course).
740 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
74210. Exim crashed if a message with an empty sender address specified by -f
743 encountered a router with an errors_to setting. This could be provoked only
744 by a command such as
746 exim -f "" ...
748 where an empty string was supplied; "<>" did not hit this bug.
75011. Installed PCRE release 4.5.
75212. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
753 remained set. It is now erased.
75513. exiqgrep wasn't working on MacOS X because it didn't correctly compute
756 times from message ids (which are base 36 rather than the normal 62).
75814. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
759 were being counted as actual protocol errors, and logged if the log
760 selector +smtp_protocol_error was set. One cannot be perfect in this test,
761 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
762 and DATA following a set of rejected RCPTs do not count as protocol errors.
763 In other words, Exim assumes they were pipelined, though this may not
764 actually be the case. Of course, in all cases the client gets an
765 appropriate error code.
76715. If a lookup fails in an ACL condition, a message about the failure may
768 be available; it is used if testing the ACL cannot continue, because most
769 such messages specify what the cause of the deferral is. However, some
770 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
771 that caused an old message to be retained and used if a later statement
772 caused a defer, replacing the real cause of the deferral.
77416. If an IP address had so many PTR records that the DNS lookup buffer
775 was not large enough to hold them, Exim could crash while trying to process
776 the truncated data. It now detects and logs this case.
77817. Further to 4.21/58, another change has been made: if (and only if) the
779 first line of a message (the first header line) ends with CRLF, a bare LF
780 in a subsequent header line has a space inserted after it, so as not to
781 terminate the header.
78318. Refactoring: tidied an ugly bit of code in appendfile that copied data
784 unnecessarily, used atoi() instead of strtol(), and didn't check the
785 termination when getting file sizes from file names by regex.
78719. Completely re-implemented the support for maildirsize files, in the light
788 of a number of problems with the previous contributed implementation
789 (4.30/29). In particular:
791 . If the quota is zero, the maildirsize file is maintained, but no quota is
792 imposed.
794 . If the maildir directory does not exist, it is created before any attempt
795 to write a maildirsize file.
797 . The quota value in the file is just a cache; if the quota is changed in
798 the transport, the new value overrides.
800 . A regular expression is available for excluding directories from the
801 count.
80320. The autoreply transport checks the characters in options that define the
804 message's headers; it allows continued headers, but it was checking with
805 isspace() after an embedded newline instead of explicitly looking for a
806 space or a tab.
80821. If all the "regular" hosts to which an address was routed had passed their
809 expiry times, and had not reached their retry times, the address was
810 bounced, even if fallback hosts were defined. Now Exim should go on to try
811 the fallback hosts.
81322. Increased buffer sizes in the callout code from 1024 to 4096 to match the
814 equivalent code in the SMTP transport. Some hosts send humungous responses
815 to HELO/EHLO, more than 1024 it seems.
81723. Refactoring: code in filter.c used (void *) for "any old type" but this
818 gives compiler warnings in some environments. I've now done it "properly",
819 using a union.
82124. The replacement for inet_ntoa() that is used with gcc on IRIX systems
822 (because of problems with the built-in one) was declared to return uschar *
823 instead of char *, causing compiler failure.
82525. Fixed a file descriptor leak when processing alias/forward files.
82726. Fixed a minor format string issue in dbfn.c.
82927. Typo in exim.c: ("dmbnz" for "dbmnz").
83128. If a filter file refered to $h_xxx or $message_headers, and the headers
832 contained RFC 2047 "words", Exim's memory could, under certain conditions,
833 become corrupted.
83529. When a sender address is verified, it is cached, to save repeating the test
836 when there is more than one recipient in a message. However, when the
837 verification involves a callout, it is possible for different callout
838 options to be set for different recipients. It is too complicated to keep
839 track of this in the cache, so now Exim always runs a verification when a
840 callout is required, relying on the callout cache for the optimization.
841 The overhead is duplication of the address routing, but this should not be
842 too great.
84430. Fixed a bug in callout caching. If a RCPT command caused the sender address
845 to be verified with callout=postmaster, and the main callout worked but the
846 postmaster check failed, the verification correctly failed. However, if a
847 subsequent RCPT command asked for sender verification *without* the
848 postmaster check, incorrect caching caused this verification also to fail,
849 incorrectly.
85131. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
852 it was not caching the DNS options (qualify_single, search_parents) that
853 were used when the lookup failed. A subsequent lookup with different
854 options therefore always gave the same answer, though there were cases
855 where it should not have. (Example: a "domains = !$mx_any" option on a
856 dnslookup router: the "domains" option is always processed without any
857 widening, but the router might have qualify_single set.) Now Exim uses the
858 cached value only when the same options are set.
86032. Added John Jetmore's "exipick" utility to the distribution.
86233. GnuTLS: When an attempt to start a TLS session fails for any reason other
863 than a timeout (e.g. a certificate is required, and is not provided), an
864 Exim server now closes the connection immediately. Previously it waited for
865 the client to close - but if the client is SSL, it seems that they each
866 wait for each other, leading to a delay before one of them times out.
86834: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
869 maintained 0.8.x compatibility because I don't think many are using it, and
870 it is clearly obsolete.
87235. Added TLS support for CRLs: a tls_crl global option and one for the smtp
873 transport.
87536. OpenSSL: $tls_certificate_verified was being set to 1 even if the
876 client certificate was expired. A simple patch fixes this, though I don't
877 understand the full logic of why the verify callback is called multiple
878 times.
88037. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
881 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
882 which causes problems with some clients (such as the Certicom SSL Plus
883 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
884 disables the coutermeasure allowing Eudora to connect."
88638. Exim was not checking that a write() to a log file succeeded. This could
887 lead to Bad Things if a log got too big, in particular if it hit a file
888 size limit. Exim now panics and dies if it cannot write to a log file, just
889 as it does if it cannot open a log file.
89139. Modified OS/Makefile-Linux so that it now contains
895 The two -D definitions ensure that Exim is compiled with large file
896 support, which makes it possible to handle log files that are bigger than
897 2^31.
89940. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
900 instance) a domain was checked against a named list that involved a lookup,
901 causing $domain_data to be set, then another domain was checked against the
902 same list, then the first domain was re-checked, the value of $domain_data
903 after the final check could be wrong. In particular, if the second check
904 failed, it could be set empty. This bug probably also applied to
905 $localpart_data.
90741. The strip_trailing_dot option was not being applied to the address given
908 with the -f command-line option.
91042. The code for reading a message's header from the spool was incrementing
911 $received_count, but never initializing it. This meant that the value was
912 incorrect (doubled) while delivering a message in the same process in which
913 it was received. In the most common configuration of Exim, this never
914 happens - a fresh exec is done - but it can happen when
915 deliver_drop_privilege is set.
91743. When Exim logs an SMTP synchronization error - client data sent too soon -
918 it now includes up to 150 characters of the unexpected data in the log
919 line.
92144. The exim_dbmbuild utility uses fixed size buffers for reading input lines
922 and building data strings. The size of both of these buffers was 10 000
923 bytes - far larger than anybody would *ever* want, thought I. Needless to
924 say, somebody hit the limit. I have increased the maximum line length to
925 20 000 and the maximum data length of concatenated lines to 100 000. I have
926 also fixed two bugs, because there was no checking on these buffers. Tsk,
927 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
928 buffer is too small.
93045. The exim_dbmbuild utility did not support quoted keys, as Exim does in
931 lsearch lookups. Now it does.
93346. When parsing a route_list item in a manualroute router, a fixed-length
934 buffer was used for the list of hosts. I made this 1024 bytes long,
935 thinking that nobody would ever have a list of hosts that long. Wrong.
936 Somebody had a whole pile of complicated expansion conditions, and the
937 string was silently truncated, leading to an expansion error. It turns out
938 that it is easier to change to an unlimited length (owing to other changes
939 that have happened since this code was originally written) than to build
940 structure for giving a limitation error. The length of the item that
941 expands into the list of hosts is now unlimited.
94347. The lsearch lookup could not handle data where the length of text line was
944 more than 4095 characters. Such lines were truncated, leading to shortened
945 data being returned. It should now handle lines of any length.
94748. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
948 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
949 ACL").
95149. Cosmetic tidy to scripts like exicyclog that are generated by globally
952 replacing strings such as BIN_DIRECTORY in a source file: the replacement
953 no longer happens in comment lines. A list of replacements is now placed
954 at the head of all of the source files, except those whose only change is
955 to replace PERL_COMMAND in the very first #! line.
95750. Replaced the slow insertion sort in queue.c, for sorting the list of
958 messages on the queue, with a bottom-up merge sort, using code contributed
959 by Michael Haardt. This should make operations like -bp somewhat faster on
960 large queues. It won't affect queue runners, except when queue_run_in_order
961 is set.
96351. Installed eximstats 1.31 in the distribution.
96552. Added support for SRV lookups to the dnslookup router.
96753. If an ACL referred to $message_body or $message_body_end, the value was not
968 reset for any messages that followed in the same SMTP session.
97054. The store-handling optimization for building very long strings was not
971 differentiating between the different store pools. I don't think this
972 actually made any difference in practice, but I've tidied it.
97455. While running the routers to verify a sender address, $sender_address
975 was still set to the sender address. This is wrong, because when routing to
976 send a bounce to the sender, it would be empty. Therefore, I have changed
977 it so that, while verifying a sender address, $sender_address is set to <>.
978 (There is no change to what happens when verifying a recipient address.)
98056. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
981 target A or AAAA records (if not already returned) without resetting the
982 qualify_single or search_parents options of the DNS resolver. These are
983 inappropriate in this case because the targets of MX and SRV records must
984 be FQDNs. A broken DNS record could cause trouble if it happened to have a
985 target that, when qualified, matched something in the local domain. These
986 two options are now turned off when doing these lookups.
98857. It seems that at least some releases of Reiserfs (which does not have the
989 concept of a fixed number of inodes) returns zero and not -1 for the
990 number of available inodes. This interacted badly with check_spool_inodes,
991 which assumed that -1 was the "no such thing" setting. What I have done is
992 to check that the total number of inodes is greater than zero before doing
993 the test of how many are available.
99558. When a "warn" ACL statement has a log_message modifier, the message is
996 remembered, and not repeated. This is to avoid a lot of repetition when a
997 message has many recipients that cause the same warning to be written.
998 Howewer, Exim was preserving the list of already written lines for an
999 entire SMTP session, which doesn't seem right. The memory is now reset if a
1000 new message is started.
100259. The "rewrite" debugging flag was not showing the result of rewriting in the
1003 debugging output unless log_rewrite was also set.
100560. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
1006 of (int)(handle) when we know that handle contains (void *)(-1).
100861. The Exim daemon panic-logs an error return when it closes the incoming
1009 connection. However "connection reset by peer" seems to be common, and
1010 isn't really an error worthy of noting specially, so that particular error
1011 is no long logged.
101362. When Exim is trying to find all the local interfaces, it used to panic and
1014 die if the ioctl to get the interface flags failed. However, it seems that
1015 on at least one OS (Solaris 9) it is possible to have an interface that is
1016 included in the list of interfaces, but for which you get a failure error
1017 for this call. This happens when the interface is not "plumbed" into a
1018 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
1019 failure of the "get flags" call assumes that the interface is down.
102163. Added a ${eval10: operator, which assumes all numbers are decimal. This
1022 makes life easier for people who are doing arithmetic on fields extracted
1023 from dates, where you often get leading zeros that should not be
1024 interpreted as octal.
102664. Added qualify_domain to the redirect router, to override the global
1027 setting.
102965. If a pathologically long header line contained very many addresses (the
1030 report of this problem mentioned 10 000) and each of them was rewritten,
1031 Exim could use up a very large amount of memory. (It kept on making new
1032 copies of the header line as it rewrote, and never released the old ones.)
1033 At the expense of a bit more processing, the header rewriting function has
1034 been changed so that it no longer eats memory in this way.
103666. The generation of the Received: header has been moved from the time that a
1037 message starts to be received, to the time that it finishes. The timestamp
1038 in the Received: header should now be very close to that of the <= log
1039 line. There are two side-effects of this change:
1041 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
1042 logged header lines no longer include the local Received: line, because
1043 it has not yet been created. The same applies to a copy of the message
1044 that is returned to a non-SMTP sender when a message is rejected.
1046 (b) When a filter file is tested using -bf, no additional Received: header
1047 is added to the test message. After some thought, I decided that this
1048 is a bug fix.
1050 This change does not affect the value of $received_for. It is still set
1051 after address rewriting, but before local_scan() is called.
105367. Installed the latest Cygwin-specific files from the Cygwin maintainer.
105568. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
1056 gave an unhelpful panic error message, and a defer error. I have managed to
1057 change this behaviour so that it now rejects any supplied certificate,
1058 which seems right, as the list of acceptable certificates is empty.
106069. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
1061 gave an unhelpful defer error. I have not managed to make this reject any
1062 supplied certificates, but the error message it gives is "no certificate
1063 supplied", which is not helpful.
106570. exigrep's output now also includes lines that are not associated with any
1066 message, but which match the given pattern. Implemented by a patch from
1067 Martin Sluka, which also tidied up the Perl a bit.
106971. Recipient callout verification, like sender verification, was using <> in
1070 the MAIL FROM command. This isn't really the right thing, since the actual
1071 sender may affect whether the remote host accepts the recipient or not. I
1072 have changed it to use the actual sender in the callout; this means that
1073 the cache record is now keyed on a recipient/sender pair, not just the
1074 recipient address. There doesn't seem to be a real danger of callout loops,
1075 since a callout by the remote host to check the sender would use <>.
1076 [SEE ABOVE: changed after hitting problems.]
107872. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
1079 temporary errors. However, in the case of such a code being given after
1080 the end of a data transmission (i.e. after ".") Exim was failing to write
1081 a retry record for the message. (Yes, there was some broken host that was
1082 actually sending 8xx at this point.)
108473. An unknown lookup type in a host list could cause Exim to panic-die when
1085 the list was checked. (An example that provoked this was putting <; in the
1086 middle of a list instead of at the start.) If this happened during a DATA
1087 ACL check, a -D file could be left lying around. This kind of configuration
1088 error no longer causes Exim to die; instead it causes a defer errror. The
1089 incident is still logged to the main and panic logs.
109174. Buglet left over from Exim 3 conversion. The message "too many messages
1092 in one connection" was written to the rejectlog but not the mainlog, except
1093 when address rewriting (yes!) was being logged.
109575. Added write_rejectlog option.
109776. When a system filter was run not as root (that is, when system_filter_user
1098 was set), the values of the $n variables were not being returned to the
1099 main process; thus, they were not subsequently available in the $sn
1100 variables.
110277. Added +return_path_on_delivery log selector.
110478. A connection timeout was being treated differently from recipients deferred
1105 when testing hosts_max_try with a message that was older than the host's
1106 retry timeout. (The host should not be counted, thus allowing all hosts to
1107 be tried at least once before bouncing.) This may have been the cause of an
1108 occasionally reported bug whereby a message would remain on the queue
1109 longer than the retry timeout, but would be bounced if a delivery was
1110 forced. I say "may" because I never totally pinned down the problem;
1111 setting up timeout/retry tests is difficult. See also the next item.
111379. The ultimate address timeout was not being applied to errors that involved
1114 a combination of host plus message (for example, a timeout on a MAIL
1115 command). When an address resolved to a number of possible hosts, and they
1116 were not all tried for each delivery (e.g. because of hosts_max_try), a
1117 message could remain on the queue longer than the retry timeout.
111980. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
1120 Haardt.
112281. Fixed an obscure SMTP outgoing bug which required at least the following
1123 conditions: (a) there was another message waiting for the same server;
1124 (b) the server returned 5xx to all RCPT commands in the first message so
1125 that the message was not completed; (c) the server dropped the connection
1126 or gave a negative response to the RSET that Exim sends to abort the
1127 transaction. The observed case was a dropped connection after DATA that had
1128 been sent in pipelining mode. That is, the server had advertised PIPELINING
1129 but was not implementing it correctly. The effect of the bug was incorrect
1130 behaviour, such as trying another host, and this could lead to a crash.
1133Exim version 4.30
1136 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
1137 and daemon.c were passed as pointers to ints; they should have been
1138 pointers to socklen_t variables (which are typically unsigned ints).
1140 2. Some signed/unsigned type warnings in the os.c file for Linux have been
1141 fixed.
1143 3. Fixed a really odd bug that affected only the testing scheme; patching a
1144 certain fixed string in the binary changed the value of another string that
1145 happened to be identical to the end of the original first string.
1147 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
1148 name", it returns that address as the IP address. On some operating
1149 systems (e.g. Solaris), it also passes back the IP address string as the
1150 "host name". However, on others (e.g. Linux), it passes back an empty
1151 string. Exim wasn't checking for this, and was changing the host name to an
1152 empty string, assuming it had been canonicized.
1154 5. Although rare, it is permitted to have more than one PTR record for a given
1155 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
1156 all the names associated with an address, because they do in Solaris.
1157 However, it seems that they do not in Linux for data that comes from the
1158 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
1159 I found this out when I moved to a new Linux workstation and tried to run
1160 the Exim test suite.
1162 To get round this problem I have changed the code so that it now does its
1163 own call to the DNS to look up PTR records when searching for a host name.
1164 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
1165 addresses that are only in /etc/hosts are still found.
1167 This behaviour is, however, controlled by an option called host_lookup_
1168 order, which defaults to "bydns:byaddr". If people want to use the other
1169 order, or indeed, just use one or the other means of lookup, they can
1170 specify it in this variable.
1172 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
1173 some operating systems, this comes back from gethostbyaddr() as an empty
1174 string, and this is what Exim used to test for. However, it seems that in
1175 other systems, "." is yielded. Exim now tests for this case too.
1177 7. The values of check_spool_space and check_log_space are now held internally
1178 as a number of kilobytes instead of an absolute number of bytes. If a
1179 numbers is specified without 'K' or 'M', it is rounded up to the nearest
1180 kilobyte. This means that much larger values can be stored.
1182 8. Exim monitor: an attempt to get the action menu when not actually pointing
1183 at a message produces an empty menu entitled "No message selected". This
1184 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
1185 no entries in it ("Shell widget menu has zero width and/or height"). So I
1186 have added a single, blank menu entry in this case.
1188 9. Added ${quote_local_part.
119010. MIME decoding is now applied to the contents of Subject: header lines when
1191 they are logged.
119311. Now that a reference to $sender_host_address automatically causes a reverse
1194 lookup to occur if necessary (4.13/18), there is no need to arrange for a
1195 host lookup before query-style lookups in lists that might use this
1196 variable. This has therefore been abolished, and the "net-" prefix is no
1197 longer necessary for query-style lookups.
119912. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
1200 have been a typo for LFLAGS, so it has been changed.
120213. The install script calls Exim with "-C /dev/null" in order to find the
1203 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
1204 to be output. Howeve, since Exim outputs its version number before the
1205 error, it didn't break the script. It just looked ugly. I fixed this by
1206 always allowing "-C /dev/null" if the caller is root.
120814. Ignore overlarge ACL variable number when reading spool file - insurance
1209 against a later release with more variables having written the file.
121115. The standard form for an IPv6 address literal was being rejected by EHLO.
1212 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1213 this, as well as the form without the "IPv6" on the front.
121516. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
1216 OS/Makefile-Darwin file.
121817. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
1219 only to LDAP libraries that do not have LDAP_OPT_DEREF.
122118. After change 4.21/52, "%ld" was used to format the contents of the $inode
1222 variable. However, some OS use ints for inodes. I've added cast to long int
1223 to get rid of the compiler warning.
122519. I had forgotten to lock out "/../" in configuration file names when
1226 ALT_CONFIG_PREFIX was set.
122820. Routers used for verification do not need to specify transports. However,
1229 if such a router generated a host list, and callout was configured, Exim
1230 crashed, because it could not find a port number from the (non-existent)
1231 transport. It now assumes port 25 in this circumstance.
123321. Added the -t option to exigrep.
123522. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
1236 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
1237 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
1238 them is set, LOOKUP_LSEARCH is forced.
124023. "exim -bV" now outputs a list of lookups that are included in the binary.
124224. Added sender and host information to the "rejected by local_scan()" log
1243 line; previously there was no indication of these.
124525. Added .include_if_exists.
124726. Change 3.952/11 added an explicit directory sync on top of a file sync for
1248 Linux. It turns out that not all file systems support this. Apparently some
1249 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
1250 do it.) To cope with this, the error EINVAL, which means that sync-ing is
1251 not supported on the file descriptor, is now ignored when Exim is trying to
1252 sync a directory. This applies only to Linux.
125427. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
125628. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
1257 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
1258 this.
126029. Added support for maildirsize files from supplied patch (modified a bit).
126230. The use of :fail: followed by an empty string could lead Exim to respond to
1263 sender verification failures with (e.g.):
1265 550 Verification failed for <xxx>
1266 550 Sender verify failed
1268 where the first response line was missing the '-' that indicates it is not
1269 the final line of the response.
127131. The loop for finding the name of the user that called Exim had a hardwired
1272 limit of 10; it now uses the value of finduser_retries, which is used for
1273 all other user lookups.
127532. Added $received_count variable, available in data and not_smtp ACLs, and at
1276 delivery time.
127833. Exim was neglecting to zero errno before one call of strtol() when
1279 expanding a string and expecting an integer value. On some systems this
1280 resulted in spurious "integer overflow" errors. Also, it was casting the
1281 result into an int without checking.
128334. Testing for a connection timeout using "timeout_connect" in the retry rules
1284 did not work. The code looks as if it has *never* worked, though it appears
1285 to have been documented since at least releast 1.62. I have made it work.
128735. The "timeout_DNS" error in retry rules, also documented since at least
1288 1.62, also never worked. As it isn't clear exactly what this means, and
1289 clearly it isn't a major issue, I have abolished the feature by treating it
1290 as "timeout", and writing a warning to the main and panic logs.
129236. The display of retry rules for -brt wasn't always showing the error code
1293 correctly.
129537. Added new error conditions to retry rules: timeout_A, timeout_MX,
1296 timeout_connect_A, timeout_connect_MX.
129838. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
1299 to the empty sender.
130139. The daemon was not analysing the content of -oX till after it had closed
1302 stderr and disconnected from the controlling terminal. This meant that any
1303 syntax errors were only noted on the panic log, and the return code from
1304 the command was 0. By re-arranging the code a little, I've made the
1305 decoding happen first, so such errors now appear on stderr, and the return
1306 code is 1. However, the actual setting up of the sockets still happens in
1307 the disconnected process, so errors there are still only recorded on the
1308 panic log.
131040. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
1311 connections (as happens on some IP stacks) was logged at start up time as
1312 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
1313 it from "IPv6 and IPv4", which means that two separate sockets are being
1314 used.
131641. The debug output for gethostbyname2() or getipnodebyname() failures now
1317 says whether AF_INET or AF_INET6 was passed as an argument.
131942. Exiwhat output was messed up when time zones were included in log
1320 timestamps.
132243. Exiwhat now gives more information about the daemon's listening ports,
1323 and whether -tls-on-connect was used.
132544. The "port" option of the smtp transport is now expanded.
132745. A "message" modifier in a "warn" statement in a non-message ACL was being
1328 silently ignored. Now an error message is written to the main and panic
1329 logs.
133146. There's a new ACL modifier called "logwrite" which writes to a log file
1332 as soon as it is encountered.
133447. Added $local_user_uid and $local_user_gid at routing time.
133648. Exim crashed when trying to verify a sender address that was being
1337 rewritten to "<>".
133949. Exim was recognizing only a space character after ".include". It now also
1340 recognizes a tab character.
134250. Fixed several bugs in the Perl script that creates the exim.8 man page by
1343 extracting the relevant information from the specification. The man page no
1344 longer contains scrambled data for the -d option, and I've added a section
1345 at the front about calling Exim under different names.
134751. Added "extra_headers" argument to the "mail" command in filter files.
134952. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
1350 crash.
135253. Installed eximstats 1.29.
135454. Added transport_filter_timeout as a generic transport option.
135655. Exim no longer adds an empty Bcc: header to messages that have no To: or
1357 Cc: header lines. This was required by RFC 822, but it not required by RFC
1358 2822.
136056. Exim used to add From:, Date:, and Message-Id: header lines to any
1361 incoming messages that did not have them. Now it does so only if the
1362 message originates locally, that is, if there is no associated remote host
1363 address. When Resent- header lines are present, this applies to the Resent-
1364 lines rather than the non-Resent- lines.
136657. Drop incoming SMTP connection after too many syntax or protocol errors. The
1367 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
136958. Messages for configuration errors now include the name of the main
1370 configuration file - useful now that there may be more than one file in a
1371 list (.included file names were always shown).
137359. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
1374 for those rare installations that do not start the daemon as root or run it
1375 setuid root. I've cut out the call to initgroups() if the daemon is not
1376 root at that time.
137860. The Exim user and group can now be bound into the binary as text strings
1379 that are looked up at the start of Exim's processing.
138161. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
138362. Added $mailstore_basename variable.
138563. Installed patch to sieve.c from Michael Haardt.
138764. When Exim failed to open the panic log after failing to open the main log,
1388 the original message it was trying to log was written to stderr and debug
1389 output, but if they were not available (the usual case in production), it
1390 was lost. Now it is written to syslog before the two lines that record the
1391 failures to open the logs.
139365. Users' Exim filters run in subprocesses under the user's uid. It is
1394 possible for a "deliver" command or an alias in a "personal" command to
1395 provoke an address rewrite. If logging of address rewriting is configured,
1396 this fails because the process is not running as root or exim. There may be
1397 a better way of dealing with this, but for the moment (because 4.30 needs
1398 to be released), I have disabled address rewrite logging when running a
1399 filter in a non-root, non-exim process.
1402Exim version 4.24
1405 1. The buildconfig auxiliary program wasn't quoting the value set for
1406 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
1407 not defined. This bug was masked in 4.22 by the effect that was fixed in
1408 change 4.23/1.
1410 2. Some messages that were rejected after a message id was allocated were
1411 shown as "incomplete" by exigrep. It no longer does this for messages that
1412 are rejected by local_scan() or the DATA or non-SMTP ACLs.
1414 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
1415 have allow_domain_literals set, the ID did not get logged in the <= line.
1416 Domain literals are now always recognized in Message-ID: header lines.
1418 4. The first argument for a ${extract expansion item is the key name or field
1419 number. Leading and trailing spaces in this item were not being ignored,
1420 causing some misleading effects.
1422 5. When deliver_drop_privilege was set, single queue runner processes started
1423 manually (i.e. by the command "exim -q") or by the daemon (which uses the
1424 same command in the process it spins off) were not dropping privilege.
1426 6. When the daemon running as "exim" started a queue runner, it always
1427 re-executed Exim in the spun-off process. This is a waste of effort when
1428 deliver_drop_privilege is set. The new process now just calls the
1429 queue-runner function directly.
1432Exim version 4.23
1435 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
1438 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
1439 ignored. Though the use of -oP was forcing the writing of a pid file, it
1440 was always written to the default place.
1442 3. If the message "no IP address found for host xxxx" is generated during
1443 incoming verification, it is now followed by identification of the incoming
1444 connection (so you can more easily find what provoked it).
1446 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
1448 5. Added some features to "harden" Exim a bit more against certain attacks:
1450 (a) There is now a build-time option called FIXED_NEVER_USERS that can
1451 be put in Local/Makefile. This is like the never_users runtime option,
1452 but it cannot be overridden. The default setting is "root".
1454 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
1455 prefix string with which any file named in a -C command line option
1456 must start.
1458 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
1459 is retained for -C and -D only if the caller of Exim is root. Without
1460 it, the exim user may also use -C and -D and retain privilege.
1462 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
1463 command line option is disabled.
1465 6. Macro names set by the -D option must start with an upper case letter, just
1466 like macro names defined in the configuration file.
1468 7. Added "dereference=" facility to LDAP.
1470 8. Two instances of the typo "uknown" in the source files are fixed.
1472 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
1473 the Configure-Makefile script screwed up while processing it.
147510. Incorporated PCRE 4.4.
147711. The SMTP synchronization check was not operating right at the start of an
1478 SMTP session. For example, it could not catch a HELO sent before the client
1479 waited for the greeting. There is now a check for outstanding input at the
1480 point when the greeting is written. Because of the duplex, asynchronous
1481 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
1482 way, but not yet received, when the check is performed.
148412. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
1485 on TCP/IP sockets, because this apparently causes some broken clients to
1486 timeout.
148813. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
1489 unchanged) from the Cygwin maintainer.
149114. The code for -bV that shows what is in the binary showed "mbx" when maildir
1492 was supported instead of testing for mbx. Effectively a typo.
149415. The spa authenticator server code was not checking that the input it
1495 received was valid base64.
149716. The debug output line for the "set" modifier in ACLs was not showing the
1498 name of the variable that was being set.
150017. Code tidy: the variable type "vtype_string" was never used. Removed it.
150218. Previously, a reference to $sender_host_name did not cause a DNS reverse
1503 lookup on its own. Something else was needed to trigger the lookup. For
1504 example, a match in host_lookup or the need for a host name in a host list.
1505 Now, if $sender_host_name is referenced and the host name has not yet been
1506 looked up, a lookup is performed. If the lookup fails, the variable remains
1507 empty, and $host_lookup_failed is set to "1".
150919. Added "eqi" as a case-independent comparison operator.
151120. The saslauthd authentication condition could segfault if neither service
1512 nor realm was specified.
151421. If an overflowing value such as "2048M" was set for message_size_limit, the
1515 error message that was logged was misleading, and incoming SMTP
1516 connections were dropped. The message is now more accurate, and temporary
1517 errors are given to SMTP connections.
151922. In some error situations (such as 21 above) Exim rejects all SMTP commands
1520 (except RSET) with a 421 error, until QUIT is received. However, it was
1521 failing to send a response to QUIT.
152323. The HELO ACL was being run before the code for helo_try_verify_hosts,
1524 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
1525 ACL is now run after the helo_try_verify_hosts code.
152724. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
1528 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
1529 case-independent, so other case variants are also recognized). Apparently
1530 some systems use these upper case variants.
153225. If more than two messages were waiting for the same host, and a transport
1533 filter was specified for the transport, Exim sent two messages over the
1534 same TCP/IP connection, and then failed with "socket operation on non-
1535 socket" when it tried to send the third.
153726. Added Exim::debug_write and Exim::log_write for embedded Perl use.
153927. The extern definition of crypt16() in expand.c was not being excluded when
1540 the OS had its own crypt16() function.
154228. Added bounce_return_body as a new option, and bounce_return_size_limit
1543 as a preferred synonym for return_size_limit, both as an option and as an
1544 expansion variable.
154629. Added LIBS=-liconv to OS/Makefile-OSF1.
154830. Changed the default configuration ACL to relax the local part checking rule
1549 for addresses that are not in any local domains. For these addresses,
1550 slashes and pipe symbols are allowed within local parts, but the sequence
1551 /../ is explicitly forbidden.
155331. SPA server authentication was not clearing the challenge buffer before
1554 using it.
155632. log_message in a "warn" ACL statement was writing to the reject log as
1557 well as to the main log, which contradicts the documentation and doesn't
1558 seem right (because no rejection is happening). So I have stopped it.
156033. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
1561 However, I am unable to do any testing of this.
156334. Fixed an infelicity in the appendfile transport. When checking directories
1564 for a mailbox, to see if any needed to be created, it was accidentally
1565 using path names with one or more superfluous leading slashes; tracing
1566 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
156835. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
1569 contents are added to the log line that is written for every discarded
1570 recipient. (Previously a log_message setting was ignored.)
157236. The ${quote: operator now quotes the string if it is empty.
157437. The install script runs exim in order to find its version number. If for
1575 some reason other than non-existence or emptiness, which it checks, it
1576 could not run './exim', it was installing it with an empty version number,
1577 i.e. as "exim-". This error state is now caught, and the installation is
1578 aborted.
158038. An argument was missing from the function that creates an error message
1581 when Exim fails to connect to the socket for saslauthd authentication.
1582 This could cause Exim to crash, or give a corrupted message.
158439. Added isip, isip4, and isip6 to ${if conditions.
158640. The ACL variables $acl_xx are now saved with the message, and can be
1587 accessed later in routers, transports, and filters.
158941. The new lookup type nwildlsearch is like wildlsearch, except that the key
1590 strings in the file are not string-expanded.
159242. If a MAIL command specified a SIZE value that was too large to fit into an
1593 int variable, the check against message_size_limit failed. Such values are
1594 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
1595 day this will have to be increased, but I don't think I want to be around
1596 when emails are that large.
1600Exim version 4.22
1603 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
1604 iconv() is not standard in FreeBSD.
1606 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
1607 IPv6 enabled. The observed symptom was a segmentation fault on return from
1608 the function os_common_find_running_interfaces() in src/os.c.
1610 3. In the check_special_case() function in daemon.c I had used "errno" as an
1611 argument name, which causes warnings on some systems. This was basically a
1612 typo, since it was named "eno" in the comments!
1614 4. The code that waits for the clock to tick (at a resolution of some fraction
1615 of a second) so as to ensure message-id uniqueness was always waiting for
1616 at least one whole tick, when it could have waited for less. [This is
1617 almost certainly not relevant at current processor speeds, where it is
1618 unlikely to ever wait at all. But we try to future-proof.]
1620 5. The function that sleeps for a time interval that includes fractions of a
1621 second contained a race. It did not block SIGALRM between setting the
1622 timer, and suspending (a couple of lines later). If the interval was short
1623 and the sigsuspend() was delayed until after it had expired, the suspension
1624 never ended. On busy systems this could lead to processes getting stuck for
1625 ever.
1627 6. Some uncommon configurations may cause a lookup to happen in a queue runner
1628 process, before it forks any delivery processes. The open lookup caching
1629 mechanism meant that the open file or database connection was passed into
1630 the delivery process. The problem was that delivery processes always tidy
1631 up cached lookup data. This could cause a problem for the next delivery
1632 process started by the queue runner, because the external queue runner
1633 process does not know about the closure. So the next delivery process
1634 still has data in the lookup cache. In the case of a file lookup, there was
1635 no problem because closing a file descriptor in a subprocess doesn't affect
1636 the parent. However, if the lookup was caching a connection to a database,
1637 the connection was closed, and the second delivery process was likely to
1638 see errors such as "PGSQL: query failed: server closed the connection
1639 unexpectedly". The problem has been fixed by closing all cached lookups
1640 in a queue runner before running a delivery process.
1642 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
1643 seem to have the "const" qualifier which it has on other OS. I've
1644 parameterised it.
1646 8. Change 4.21/2 was too strict. It is only if there are two authenticators
1647 *of the same type* (client or server) with the same public name that an
1648 error should be diagnosed.
1650 9. When Exim looked up a host name for an IP address, but failed to find the
1651 original IP address when looking up the host name (a safety check), it
1652 output the message "<ip address> does not match any IP for NULL", which was
1653 confusing, to say the least. The bug was that the host name should have
1654 appeared instead of "NULL".
165610. Since release 3.03, if Exim is called by a uid other than root or the Exim
1657 user that is built into the binary, and the -C or -D options is used, root
1658 privilege is dropped before the configuration file is read. In addition,
1659 logging is switched to stderr instead of the normal log files. If the
1660 configuration then re-defines the Exim user, the unprivileged environment
1661 is probably not what is expected, so Exim logs a panic warning message (but
1662 proceeds).
1664 However, if deliver_drop_privilege is set, the unprivileged state may well
1665 be exactly what is intended, so the warning has been cut out in that case,
1666 and Exim is allowed to try to write to its normal log files.
1669Exim version 4.21
1672 1. smtp_return_error_details was not giving details for temporary sender
1673 or receiver verification errors.
1675 2. Diagnose a configuration error if two authenticators have the same public
1676 name.
1678 3. Exim used not to create the message log file for a message until the first
1679 delivery attempt. This could be confusing when incoming messages were held
1680 for policy or load reasons. The message log file is now created at the time
1681 the message is received, and an initial "Received" line is written to it.
1683 4. The automatically generated man page for command line options had a minor
1684 bug that caused no ill effects; however, a more serious problem was that
1685 the procedure for building the man page automatically didn't always
1686 operate. Consequently, release 4.20 contains an out-of-date version. This
1687 shouldn't happen again.
1689 5. When building Exim with embedded Perl support, the script that builds the
1690 Makefile was calling 'perl' to find its compile-time parameters, ignoring
1691 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
1693 6. The freeze_tell option was not being used for messages that were frozen on
1694 arrival, either by an ACL or by local_scan().
1696 7. Added the smtp_incomplete_transaction log selector.
1698 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
1699 was accepting AUTH without a new EHLO.
1701 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
1702 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
1703 RFC.
170510. Logging of TCP/IP connections (when configured) now happens in the main
1706 daemon process instead of the child process, so that the TCP/IP connection
1707 count is more accurate (but it can never be perfect).
170911. The use of "drop" in a nested ACL was not being handled correctly in the
1710 outer ACL. Now, if condition failure induced by the nested "drop" causes
1711 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
1712 or "require"), the connection is dropped.
171412. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
1715 that yield "discard" can now be used with an "accept" or a "discard" verb,
1716 but an error is generated for any others (because I can't see a useful way
1717 to define what should happen).
171913. When an ACL is read dynamically from a file (or anywhere else), the lines
1720 are now processed in the same way as lines in the Exim configuration file.
1721 In particular, continuation lines are supported.
172314. Added the "dnslists = a.b.c!=n.n.n.n" feature.
172515. Added -ti meaning -t -i.
172716. Check for letters, digits, hyphens, and dots in the names of dnslist
1728 domains, and warn by logging if others are found.
173017. At least on BSD, alignment is not guarenteed for the array of ifreq's
1731 returned from GIFCONF when Exim is trying to find the list of interfaces on
1732 a host. The code in os.c has been modified to copy each ifreq to an aligned
1733 structure in all cases.
1735 Also, in some cases, the returned ifreq's were being copied to a 'struct
1736 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
1737 means the last couple of bytes of an IPv6 address could be chopped if the
1738 ifreq contained only a normal sockaddr (14 bytes storage).
174018. Named domain lists were not supported in the hosts_treat_as_local option.
1741 An entry such as +xxxx was not recognized, and was treated as a literal
1742 domain name.
174419. Ensure that header lines added by a DATA ACL are included in the reject log
1745 if the ACL subsequently rejects the message.
174720. Upgrade the utility script to use Digest::MD5 instead of just
1748 MD5 (which is deprecated).
175021. When testing a filter file using -bf, Exim was writing a message when it
1751 took the sender from a "From " line in the message, but it was not doing so
1752 when it took $return_path from a Return-Path: header line. It now does.
175422. If the contents of a "message" modifier for a "warn" ACL verb do not begin
1755 with a valid header line field name (a series of printing characters
1756 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
175823. Changed "disc" in the source to "disk" to conform to the documentation and
1759 the book and for uniformity.
176124. Ignore Sendmail's -Ooption=value command line item.
176325. When execve() failed while trying to run a command in a pipe transport,
1764 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
1765 could be confused with a return value of 69 from the command itself. This
1766 has been changed to 127, the value the shell returns if it is asked to run
1767 a non-existent command. The wording for the related log line suggests a
1768 non-existent command as the problem.
177026. If received_header_text expands to an empty string, do not add a Received:
1771 header line to the message. (Well, it adds a token one on the spool, but
1772 marks it "old" so that it doesn't get used or transmitted.)
177427. Installed eximstats 1.28 (addition of -nt option).
177628. There was no check for failure on the call to getsockname() in the daemon
1777 code. This can fail if there is a shortage of resources on the system, with
1778 ENOMEM, for example. A temporary error is now given on failure.
178029. Contrary to the C standard, it seems that in some environments, the
1781 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
1782 program. Exim now does this explicitly; it affects the formatting of
1783 timestamps using strftime().
178530. If exiqsumm was given junk data, it threw up some uninitialized variable
1786 complaints. I've now initialized all the variables, to avoid this.
178832. Header lines added by a system filter were not being "seen" during
1789 transport-time rewrites.
179133. The info_callback() function passed to OpenSSL is set up with type void
1792 (*)(SSL *, int, int), as described somewhere. However, when calling the
1793 function (actually a macro) that sets it up, the type void(*)() is
1794 expected. I've put in a cast to prevent warnings from picky compilers.
179634. If a DNS black list lookup found a CNAME record, but there were no A
1797 records associated with the domain it pointed at, Exim crashed.
179935. If a DNS black list lookup returned more than one A record, Exim ignored
1800 all but the first. It now scans all returned addresses if a particular IP
1801 value is being sought. In this situation, the contents of the
1802 $dnslist_value variable are a list of all the addresses, separated by a
1803 comma and a space.
180536. Tightened up the rules for host name lookups using reverse DNS. Exim used
1806 to accept a host name and all its aliases if the forward lookup for any of
1807 them yielded the IP address of the incoming connection. Now it accepts only
1808 those names whose forward lookup yields the correct IP address. Any other
1809 names are discarded. This closes a loophole whereby a rogue DNS
1810 administrator could create reverse DNS records to break through a
1811 wildcarded host restriction in an ACL.
181337. If a user filter or a system filter that ran in a subprocess used any of
1814 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
1815 the wrong values were passed to the pipe command ($thisaddress had the
1816 value of $0, $0 had the value of $1, etc). This bug was introduced by
1817 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
181938. Improved the line breaking for long SMTP error messages from ACLs.
1820 Previously, if there was no break point between 40 and 75 characters, Exim
1821 left the rest of the message alone. Two changes have been made: (a) I've
1822 reduced the minimum length to 35 characters; (b) if it can't find a break
1823 point between 35 and 75 characters, it looks ahead and uses the first one
1824 that it finds. This may give the occasional overlong line, but at least the
1825 remaining text gets split now.
182739. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
1828 file descriptors might be low, and that setting 1000 would always raise it.
1829 It turns out that in some environments, the limit is already over 1000 and
1830 that lowering it causes trouble. So now Exim takes care not to decrease it.
183240. When delivering a message, the value of $return_path is set to $sender_
1833 address at the start of routing (routers may change the value). By an
1834 oversight, this default was not being set up when an address was tested by
1835 -bt or -bv, which affected the outcome if any router or filter referred to
1836 $return_path.
183841. The idea of the "warn" ACL verb is that it adds a header or writes to the
1839 log only when "message" or "log_message" are set. However, if one of the
1840 conditions was an address verification, or a call to a nested ACL, the
1841 messages generated by the underlying test were being passed through. This
1842 no longer happens. The underlying message is available in $acl_verify_
1843 message for both "message" and "log_message" expansions, so it can be
1844 passed through if needed.
184642. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
1847 new expansion $bh_ to give the encoded byte string without charset
1848 translation. Translation happens only if iconv() is available; HAVE_ICONV
1849 indicates this at build time. HEADERS_CHARSET gives the charset to
1850 translate to; headers_charset can change it in the configuration, and
1851 "headers charset" can change it in an individual filter file.
185343. Now that we have a default RFC 2047 charset (see above), the code in Exim
1854 that creates RFC 2047 encoded "words" labels them as that charset instead
1855 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
1856 expansion operator; (ii) when Exim creates a From: line for a local
1857 message; (iii) when a header line is rewritten to include a "phrase" part.
185944. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
1860 buggy, causing it to skip the first lines of messages whose message ID
1861 ended in 'D'. This would not have bitten before Exim release 4.14, because
1862 message IDs were unlikely to end in 'D' before then. The effect was to have
1863 incorrect size information for certain domains.
186545. #include "config.h" was missing at the start of the crypt16.c module. This
1866 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
1867 noticed.
186946. If there was a timeout during a "random" callout check, Exim treated it as
1870 a failure of the random address, and carried on sending RSET and the real
1871 address. If the delay was just some slowness somewhere, the response to the
1872 original RCPT would be taken as a response to RSET and so on, causing
1873 mayhem of various kinds.
187547. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
1876 when I implemented it. It didn't allow for the fact that some option values
1877 may legitimatetly be negative (e.g. size_addition), and it didn't even do
1878 the right test for positive values.
188048. Domain names in DNS records are case-independent. Exim always looks them up
1881 in lower case. Some resolvers return domain names in exactly the case they
1882 appear in the zone file, that is, they may contain uppercase letters. Not
1883 all resolvers do this - some return always lower case. Exim was treating a
1884 change of case by a resolver as a change of domain, similar to a widening
1885 of a domain abbreviation. This triggered its re-routing code and so it was
1886 trying to route what was effectively the same domain again. This normally
1887 caused routing to fail (because the router wouldn't handle the domain
1888 twice). Now Exim checks for this case specially, and just changes the
1889 casing of the domain that it ultimately uses when it transmits the message
1890 envelope.
189249. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
1893 module.
189550. If a filter generated a file delivery with a non-absolute name (possible if
1896 no home directory exists for the router), the forbid_file option was not
1897 forbidding it.
189951. Added '&' feature to dnslists, to provide bit mask matching in addition to
1900 the existing equality matching.
190252. Exim was using ints instead of ino_t variables in some places where it was
1903 dealing with inode numbers.
190553. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
1906 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
1907 called TMPDIR, and if it finds it is different, it changes its value.
190954. The smtp_printf() function is now made available to local_scan() so
1910 additional output lines can be written before returning. There is also an
1911 smtp_fflush() function to enable the detection of a dropped connection.
1912 The variables smtp_input and smtp_batched_input are exported to
1913 local_scan().
191555. Changed the default runtime configuration: the message "Unknown user"
1916 has been removed from the ACL, and instead placed on the localuser router,
1917 using the cannot_route_message feature. This means that any verification
1918 failures that generate their own messages won't get overridden. Similarly,
1919 the "Unrouteable address" message that was in the ACL for unverifiable
1920 relay addresses has also been removed.
192256. Added hosts_avoid_esmtp to the smtp transport.
192457. The exicyclog script was not checking for the esoteric option
1925 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
1926 will work only if exicyclog is run under the appropriate euid.
192858. Following a discussion on the list, the rules by which Exim recognises line
1929 endings on incoming messages have been changed. The -dropcr and drop_cr
1930 options are now no-ops, retained only for backwards compatibility. The
1931 following line terminators are recognized: LF CRLF CR. However, special
1932 processing applies to CR:
1934 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
1935 nor a local message in the state where . is a terminator.
1937 (ii) If a bare CR is encountered in a header line, an extra space is added
1938 after the line terminator so as not to end the header. The reasoning
1939 behind this is that bare CRs in header lines are most likely either
1940 to be mistakes, or people trying to play silly games.
194259. The size of a message, as listed by "-bp" or in the Exim monitor window,
1943 was being incorrectly given as 18 bytes larger than it should have been.
1944 This is a VOB (very old bug).
194660. This may never have affected anything current, but just in case it has:
1947 When the local host is found other than at the start of a list of hosts,
1948 the local host, those with the same MX, and any that follow, are discarded.
1949 When the list in question was part of a longer list of hosts, the following
1950 hosts (not currently being processed) were also being discarded. This no
1951 longer happens. I'm not sure if this situation could ever has previously
1952 arisen.
195461. Added the "/MX" feature to lists of hosts in the manualroute and query
1955 program routers.
195762. Whenever Exim generates a new message, it now adds an Auto-Submitted:
1958 header. This is something that is recommended in a new Internet Draft, and
1959 is something that is documented as being done by Sendmail. There are two
1960 possible values. For messages generated by the autoreply transport, Exim
1961 adds:
1963 Auto-Submitted: auto-replied
1965 whereas for all other generated messages (e.g. bounces) it adds
1967 Auto-Submitted: auto-generated
196963. The "personal" condition in filters now includes a test for the
1970 Auto-Submitted: header. If it contains the string "auto-" the message it
1971 not considered personal.
197364. Added rcpt_include_affixes as a generic transport option.
197565. Added queue_only_override (default true).
197766. Added the syslog_duplication option.
197967. If what should have been the first header line of a message consisted of
1980 a space followed by a colon, Exim was mis-interpreting it as a header line.
1981 It isn't of course - it is syntactically invalid and should therefore be
1982 treated as the start of the message body. The misbehaviour could have
1983 caused a number of strange effects, including loss of data in subsequent
1984 header lines, and spool format errors.
198668. Formerly, the AUTH parameter on a MAIL command was trusted only if the
1987 client host had authenticated. This control can now be exercised by an ACL
1988 for more flexibility.
199069. By default, callouts do not happen when testing with -bh. There is now a
1991 variant, -bhc, which does actually run the callout code, including
1992 consulting and updating the callout cache.
199470. Added support for saslauthd authentication, courtesy of Alexander
1995 Sabourenkov.
199771. If statvfs() failed on the spool or log directories while checking their
1998 size for availability, Exim confusingly gave the error "space shortage".
1999 Furthermore, in debugging mode it crashed with a floating point exception.
2000 These checks are done if check_{spool,log}_{space,inodes} are set, and when
2001 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
2002 serious problem, Exim now writes to the main and panic logs when this
2003 happens, with details of the failure. It then refuses to accept the
2004 incoming message, giving the message "spool directory problem" or "log
2005 directory problem" with a 421 code for SMTP messages.
200772. When Exim is about to re-exec itself, it ensures that the file descriptors
2008 0, 1, and 2 exist, because some OS complain for execs without them (see
2009 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
2010 descriptors. However, the code omitted to check that the open succeeded,
2011 causing mysterious errors if for some reason the permissions on /dev/null
2012 got screwed. Now Exim writes a message to the main and panic logs, and
2013 bombs out if it can't open /dev/null.
201573. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
2016 interact so that it is all more flexible. It is supposed to remain
2017 backwards compatible. Also added extra_local_interfaces.
201974. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
2020 to bomb out with an assertion failure - to the client this appears as a
2021 connection drop. This problem occurs in the part of the code that was taken
2022 from the Samba project. Fortunately, the assertion is in a very simple
2023 function, so I have fixed this by reproducing the function inline in the
2024 one place where it is called, and arranging for authentication to fail
2025 instead of killing the process with assert().
202775. The SPA client code was not working when the server requested OEM rather
2028 than Unicode encoding.
203076. Added code to make require_files with a specific uid setting more usable in
2031 the case where statting the file as root fails - usually a non-root-mounted
2032 NFS file system. When this happens and the failure is EACCES, Exim now
2033 forks a subprocess and does the per-uid checking as the relevant uid.
203577. Added process_log_path.
203778. If log_file_path was not explicitly set, a setting of check_log_space or
2038 check_log_inodes was ignored.
204079. If a space check for the spool or log partitions fails, the incident is now
2041 logged. Of course, in the latter case the data may get lost...
204380. Added the %p formatting code to string_format() so that it can be used to
2044 print addresses in debug_print(). Adjusted all the address printing in the
2045 debugging in store.c to use %p rather than %d.
204781. There was a concern that a line of code in smtp_in.c could overflow a
2048 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
2049 initially expressed, the concern was not well-founded, because trailing
2050 spaces are removed early. However, if the trailing spaces were followed by
2051 a NULL, they did not get removed, so the overflow was possible. Two fixes
2052 were applied:
2054 (a) I re-wrote the offending code in a cleaner fashion.
2055 (b) If an incoming SMTP command contains a NULL character, it is rejected
2056 as invalid.
205882. When Exim changes uid/gid to the Exim user at daemon start time, it now
2059 runs initgroups(), so that if the Exim user is in any additional groups,
2060 they will be used during message reception.
2063Exim version 4.20
2066The change log for 4.20 and earlier releases has been archived.