1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.19 2004/11/10 10:29:56 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
10 1. Minor wording change to the doc/README.SIEVE file.
12 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
13 computation of the current number of files was incorrect.
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
19 4. Give more explanation in the error message when the command for a transport
20 filter fails to execute.
22 5. There are several places where Exim runs a non-Exim command in a
23 subprocess. The SIGUSR1 signal should be disabled for these processes. This
24 was being done only for the command run by the queryprogram router. It is
25 now done for all such subprocesses. The other cases are: ${run, transport
26 filters, and the commands run by the lmtp and pipe transports.
28 6. Added CONFIGURE_GROUP build-time option.
30 7. Some older OS have a limit of 256 on the maximum number of file
31 descriptors. Exim was using setrlimit() to set 1000 as a large value
32 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
33 systems. I've change it so that if it can't get 1000, it tries for 256.
35 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
36 was an oversight, and furthermore, ever since the addition of extra
37 controls (e.g. 4.43/32), the checks on when to allow different forms of
38 "control" were broken. There should now be diagnostics for all cases when a
39 control that does not make sense is encountered.
41 9. Added the /retain_sender option to "control=submission".
43 10. $recipients is now available in the predata ACL (oversight).
45 11. Tidy the search cache before the fork to do a delivery from a message
46 received from the command line. Otherwise the child will trigger a lookup
47 failure and thereby defer the delivery if it tries to use (for example) a
48 cached ldap connection that the parent has called unbind on.
50 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
51 of $address_data from the recipient verification was clobbered by the
54 13. The value of address_data from a sender verification is now available in
55 $sender_address_data in subsequent conditions in the ACL statement.
57 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
59 15. Added a new option "connect=<time>" to callout options, to set a different
62 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
63 was its contents. (It was OK if the option was not defined at all.)
65 17. A "Completed" log line is now written for messages that are removed from
66 the spool by the -Mrm option.
68 18. New variables $sender_verify_failure and $recipient_verify_failure contain
69 information about exactly what failed.
71 19. Added -dd to debug only the daemon process.
77 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
78 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
79 Now it returns an integer. A number of calls were still expecting a T/F
80 return. Fortuitously, in all cases, the tests worked in OK situations,
81 which is the norm. However, things would have gone wrong on any write
82 failures on the smtp file descriptor. This function is used when sending
83 messages over SMTP and also when doing verify callouts.
85 2. When Exim is called to do synchronous delivery of a locally submitted
86 message (the -odf or -odi options), it no longer closes stderr before doing
89 3. Implemented the mua_wrapper option.
91 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
93 5. Implemented the functions header_remove(), header_testname(),
94 header_add_at_position(), and receive_remove_recipient(), and exported them
97 6. If an ACL "warn" statement specified the addition of headers, Exim already
98 inserted X-ACL-Warn: at the start if there was no header name. However, it
99 was not making this test for the second and subsequent header lines if
100 there were newlines in the string. This meant that an invalid header could
101 be inserted if Exim was badly configured.
103 7. Allow an ACL "warn" statement to add header lines at the start or after all
104 the Received: headers, as well as at the end.
106 8. Added the rcpt_4xx retry error code.
108 9. Added postmaster_mailfrom=xxx to callout verification option.
110 10. Added mailfrom=xxxx to the callout verification option, for verify=
113 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
114 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
116 12. Inserted some casts to stop certain compilers warning when using pointer
117 differences as field lengths or precisions in printf-type calls (mostly
118 affecting debugging statements).
120 13. Added optional readline() support for -be (dynamically loaded).
122 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
123 same clock tick as a message's arrival, so that its received time was the
124 same as the "first fail" time on the retry record, and that message
125 remained on the queue past the ultimate address timeout, every queue runner
126 would try a delivery (because it was past the ultimate address timeout) but
127 after another failure, the ultimate address timeout, which should have then
128 bounced the address, did not kick in. This was a "< instead of <=" error;
129 in most cases the first failure would have been in the next clock tick
130 after the received time, and all would be well.
132 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
133 being recognized when the domain list was tested by the match_domain
134 condition in an expansion string.
136 16. Added the ${str2b64: operator.
138 17. Exim was always calling setrlimit() to set a large limit for the number of
139 processes, without checking whether the existing limit was already
140 adequate. (It did check for the limit on file descriptors.) Furthermore,
141 errors from getrlimit() and setrlimit() were being ignored. Now they are
142 logged to the main and panic logs, but Exim does carry on, to try to do its
143 job under whatever limits there are.
145 18. Imported PCRE 5.0.
147 19. Trivial typo in log message " temporarily refused connection" (the leading
150 20. If the log selector return_path_on_delivery was set and an address was
151 redirected to /dev/null, the delivery process crashed because it assumed
152 that a return path would always be set for a "successful" delivery. In this
153 case, the whole delivery is bypassed as an optimization, and therefore no
156 21. Internal re-arrangement: the function for sending a challenge and reading
157 a response while authentication was assuming a zero-terminated challenge
158 string. It's now changed to take a pointer and a length, to allow for
159 binary data in such strings.
161 22. Added the cyrus_sasl authenticator (code supplied by MBM).
163 23. Exim was not respecting finduser_retries when seeking the login of the
164 uid under which it was called; it was always trying 10 times. (The default
165 setting of finduser_retries is zero.) Also, it was sleeping after the final
166 failure, which is pointless.
168 24. Implemented tls_on_connect_ports.
170 25. Implemented acl_smtp_predata.
172 26. If the domain in control=submission is set empty, Exim assumes that the
173 authenticated id is a complete email address when it generates From: or
174 Sender: header lines.
176 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
177 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
178 chown and chgrp in /bin and hostname in /usr/bin.
180 28. Exim was keeping the "process log" file open after each use, just as it
181 does for the main log. This opens the possibility of it remaining open for
182 long periods when the USR1 signal hits a daemon. Occasional processlog
183 errors were reported, that could have been caused by this. Anyway, it seems
184 much more sensible not to leave this file open at all, so that is what now
187 29. The long-running daemon process does not normally write to the log once it
188 has entered its main loop, and it closes the log before doing so. This is
189 so that log files can straightforwardly be renamed and moved. However,
190 there are a couple of unusual error situations where the daemon does write
191 log entries, and I had neglected to close the log afterwards.
193 30. The text of an SMTP error response that was received during a remote
194 delivery was being truncated at 512 bytes. This is too short for some of
195 the long messages that one sometimes sees. I've increased the limit to
198 31. It is now possible to make retry rules that apply only when a message has a
199 specific sender, in particular, an empty sender.
201 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
202 it possible to be selective about when SMTP synchronization is enforced.
204 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
206 32. Implemented hosts_connection_nolog.
208 33. Added an ACL for QUIT.
210 34. Setting "delay_warning=" to disable warnings was not working; it gave a
213 35. Added mailbox_size and mailbox_filecount to appendfile.
215 36. Added control = no_multiline_responses to ACLs.
217 37. There was a bug in the logic of the code that waits for the clock to tick
218 in the case where the clock went backwards by a substantial amount such
219 that the microsecond fraction of "now" was more than the microsecond
220 fraction of "then" (but the whole seconds number was less).
222 38. Added support for the libradius Radius client library this is found on
223 FreeBSD (previously only the radiusclient library was supported).
229 1. When certain lookups returned multiple values in the form name=value, the
230 quoting of the values was not always being done properly. Specifically:
231 (a) If the value started with a double quote, but contained no whitespace,
233 (b) If the value contained whitespace other than a space character (i.e.
234 tabs or newlines or carriage returns) it was not quoted.
235 This fix has been applied to the mysql and pgsql lookups by writing a
236 separate quoting function and calling it from the lookup code. The fix
237 should probably also be applied to nisplus, ibase and oracle lookups, but
238 since I cannot test any of those, I have not disturbed their existing code.
240 2. A hit in the callout cache for a specific address caused a log line with no
241 reason for rejecting RCPT. Now it says "Previous (cached) callout
242 verification failure".
244 3. There was an off-by-one bug in the queryprogram router. An over-long
245 return line was truncated at 256 instead of 255 characters, thereby
246 overflowing its buffer with the terminating zero. As well as fixing this, I
247 have increased the buffer size to 1024 (and made a note to document this).
249 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
250 when Exim is waiting for an SMTP response from a remote server, Exim
251 restarts its select() call on the socket, thereby resetting its timeout.
252 This is not a problem when such interrupts are rare. Somebody set up a cron
253 job to run exiwhat every 2 minutes, which is less than the normal select()
254 timeout (5 or 10 minutes). This meant that the select() timeout never
255 kicked in because it was always reset. I have fixed this by comparing the
256 time when an interrupt arrives with the time at the start of the first call
257 to select(). If more time than the timeout has elapsed, the interrupt is
258 treated as a timeout.
260 5. Some internal re-factoring in preparation for the addition of Sieve
261 extensions (by MH). In particular, the "personal" test is moved to a
262 separate function, and given an option for scanning Cc: and Bcc: (which is
263 not set for Exim filters).
265 6. When Exim created an email address using the login of the caller as the
266 local part (e.g. when creating a From: or Sender: header line), it was not
267 quoting the local part when it contained special characters such as @.
269 7. Installed new OpenBSD configuration files.
271 8. Reworded some messages for syntax errors in "and" and "or" conditions to
272 try to make them clearer.
274 9. Callout options, other than the timeout value, were being ignored when
275 verifying sender addresses in header lines. For example, when using
277 verify = header_sender/callout=no_cache
279 the cache was (incorrectly) being used.
281 10. Added a missing instance of ${EXE} to the exim_install script; this affects
282 only the Cygwin environment.
284 11. When return_path_on_delivery was set as a log selector, if different remote
285 addresses in the same message used different return paths and parallel
286 remote delivery occurred, the wrong values would sometimes be logged.
287 (Whenever a remote delivery process finished, the return path value from
288 the most recently started remote delivery process was logged.)
290 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
291 lines when AUTH and/or TLS are in use. This is the "received protocol"
292 field. Exim used to use "asmtp" for authenticated SMTP, without any
293 indication (in the protocol name) for TLS use. Now it follows the RFC and
294 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
295 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
296 names appear in log lines as well as in Received: header lines.
298 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
299 extensions, and comparison tests, and to fix some bugs.
301 14. Changes to the "personal" filter test:
303 (1) The test was buggy in that it was just doing the equivalent of
304 "contains" tests on header lines. For example, if a user's address was
305 anne@some.where, the "personal" test would incorrectly be true for
307 To: susanne@some.where
309 This test is now done by extracting each address from the header in turn,
310 and checking the entire address. Other tests that are part of "personal"
311 are now done using regular expressions (for example, to check local parts
312 of addresses in From: header lines).
314 (2) The list of non-personal local parts in From: addresses has been
315 extended to include "listserv", "majordomo", "*-request", and "owner-*",
316 taken from the Sieve specification recommendations.
318 (3) If the message contains any header line starting with "List-" it is
319 treated as non-personal.
321 (4) The test for "circular" in the Subject: header line has been removed
322 because it now seems ill-conceived.
324 15. Minor typos in src/EDITME comments corrected.
326 16. Installed latest exipick from John Jetmore.
328 17. If headers_add on a router specified a text string that was too long for
329 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
330 of string_sprintf() is now avoided.
332 18. $message_body_size was not set (it was always zero) when running the DATA
333 ACL and the local_scan() function.
335 19. For the "mail" command in an Exim filter, no default was being set for
336 the once_repeat time, causing a random time value to be used if "once" was
337 specified. (If the value happened to be <= 0, no repeat happened.) The
338 default is now 0s, meaning "never repeat". The "vacation" command was OK
339 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
340 (I found it when inspecting the code).
342 20. There is now an overall timeout for performing a callout verification. It
343 defaults to 4 times the callout timeout, which applies to individual SMTP
344 commands during the callout. The overall timeout applies when there is more
345 than one host that can be tried. The timeout is checked before trying the
346 next host. This prevents very long delays if there are a large number of
347 hosts and all are timing out (e.g. when the network connections are timing
348 out). The value of the overall timeout can be changed by specifying an
349 additional sub-option for "callout", called "maxwait". For example:
351 verify = sender/callout=5s,maxwait=20s
353 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
354 to the end before writing, but this should make it even safer).
356 22. Exim was forgetting that it had advertised PIPELINING for the second and
357 subsequent messages on an SMTP connection. It was also not resetting its
358 memory on STARTTLS and an internal HELO.
360 23. When Exim logs an SMTP synchronization error within a session, it now
361 records whether PIPELINING has been advertised or not.
363 24. Added 3 instances of "(long int)" casts to time_t variables that were being
364 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
365 rather than long int.
367 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
369 26. Added the never_mail option to autoreply.
375 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
376 crash if the getsockname() call failed; this can happen if a connection is
377 closed very soon after it is established. The problem was simply in the
378 order in which certain operations were done, causing Exim to try to write
379 to the SMTP stream before it had set up the file descriptor. The bug has
380 been fixed by making things happen in the correct order.
386 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
387 before the connection was closed, thus losing the rejection response.
389 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
390 some early Solaris releases, but causes trouble in current releases where
391 socklen_t is defined.
393 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
396 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
399 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
400 that releasing the top of it at the end releases what was used for sub-
401 expansions (unless the block got too big). However, discard this block if
402 the first thing is a variable or header, so that we can use its block when
403 it is dynamic (useful for very large $message_headers, for example).
405 6. Lookups now cache *every* query, not just the most recent. A new, separate
406 store pool is used for this. It can be recovered when all lookup caches are
407 flushed. Lookups now release memory at the end of their result strings.
408 This has involved some general refactoring of the lookup sources.
410 7. Some code has been added to the store_xxx() functions to reduce the amount
411 of flapping under certain conditions.
413 8. log_incoming_interface used to affect only the <= reception log lines. Now
414 it causes the local interface and port to be added to several more SMTP log
415 lines, for example "SMTP connection from", and rejection lines.
417 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
419 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
421 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
422 could be overwritten at the end of the current message (or the start of a
423 new message if it was set in a HELO ACL). The value is now preserved for
424 the duration of the SMTP connection.
426 12. If a transport had a headers_rewrite setting, and a matching header line
427 contained an unqualified address, that address was qualified, even if it
428 did not match any rewriting rules. The underlying bug was that the values
429 of the flags that permit the existence of unqualified sender and recipient
430 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
431 non-local messages, and by -bnq for local messages) were not being
432 preserved with the message after it was received.
434 13. When Exim was logging an SMTP synchronization error, it could sometimes log
435 "next input=" as part of the text comprising the host identity instead of
436 the correct text. The code was using the same buffer for two different
437 strings. However, depending on which order the printing function evaluated
438 its arguments, the bug did not always show up. Under Linux, for example, my
439 test suite worked just fine.
441 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
442 doesn't work with some older versions of Perl. It has been changed to "my",
443 which in any case is probably the better facility to use.
445 15. A really picky compiler found some instances of statements for creating
446 error messages that either had too many or two few arguments for the format
449 16. The size of the buffer for calls to the DNS resolver has been increased
450 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
451 for addresses that have a lot of PTR records. This alleviates a problem; it
452 does not fully solve it.
454 17. A dnsdb lookup for PTR records that receives more data than will fit in the
455 buffer now truncates the list and logs the incident, which is the same
456 action as happens when Exim is looking up a host name and its aliases.
457 Previously in this situation something unpredictable would happen;
458 sometimes it was "internal error: store_reset failed".
460 18. If a server dropped the connection unexpectedly when an Exim client was
461 using GnuTLS and trying to read a response, the client delivery process
462 crashed while trying to generate an error log message.
464 19. If a "warn" verb in an ACL added multiple headers to a message in a single
467 warn message = H1: something\nH2: something
469 the text was added as a single header line from Exim's point of view
470 though it ended up OK in the delivered message. However, searching for the
471 second and subsequent header lines using $h_h2: did not work. This has been
472 fixed. Similarly, if a system filter added multiple headers in this way,
473 the routers could not see them.
475 20. Expanded the error message when iplsearch is called with an invalid key to
476 suggest using net-iplsearch in a host list.
478 21. When running tests using -bh, any delays imposed by "delay" modifiers in
479 ACLs are no longer actually imposed (and a message to that effect is
482 22. If a "gecos" field in a passwd entry contained escaped characters, in
483 particular, if it contained a \" sequence, Exim got it wrong when building
484 a From: or a Sender: header from that name. A second bug also caused
485 incorrect handling when an unquoted " was present following a character
488 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
489 was not being matched caselessly.
491 24. Arranged for all hyphens in the exim.8 source to be escaped with
494 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
495 properly. Recipient callout cache records were still being keyed to include
496 the sender, even when use_sender was set false. This led to far more
497 callouts that were necessary. The sender is no longer included in the key
498 when use_sender is false.
500 26. Added "control = submission" modifier to ACLs.
502 27. Added the ${base62d: operator to decode base 62 numbers.
504 28. dnsdb lookups can now access SRV records.
506 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
507 the configuration file.
509 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
510 (-v) mode. This makes the output for a verbose queue run more intelligible.
512 31. Added a use_postmaster feature to recipient callouts.
514 32. Added the $body_zerocount variable, containing the number of binary zero
515 bytes in the message body.
517 33. The time of last modification of the "new" subdirectory is now used as the
518 "mailbox time last read" when there is a quota error for a maildir
521 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
523 35. Added +ignore_unknown as a special item in host lists.
525 36. Code for decoding IPv6 addresses in host lists is now included, even if
526 IPv6 support is not being compiled. This fixes a bug in which an IPv6
527 address was recognized as an IP address, but was then not correctly decoded
528 into binary, causing unexpected and incorrect effects when compared with
535 1. Very minor rewording of debugging text in manualroute to say "list of
536 hosts" instead of "hostlist".
538 2. If verify=header_syntax was set, and a header line with an unqualified
539 address (no domain) and a large number of spaces between the end of the
540 name and the colon was received, the reception process suffered a buffer
541 overflow, and (when I tested it) crashed. This was caused by some obsolete
542 code that should have been removed. The fix is to remove it!
544 3. When running in the test harness, delay a bit after writing a bounce
545 message to get a bit more predictability in the log output.
547 4. Added a call to search_tidyup() just before forking a reception process. In
548 theory, someone could use a lookup in the expansion of smtp_accept_max_
549 per_host which, without the tidyup, could leave open a database connection.
551 5. Added the variables $recipient_data and $sender_data which get set from a
552 lookup success in an ACL "recipients" or "senders" condition, or a router
553 "senders" option, similar to $domain_data and $local_part_data.
555 6. Moved the writing of debug_print from before to after the "senders" test
558 7. Change 4.31/66 (moving the time when the Received: is generated) caused
559 problems for message scanning, either using a data ACL, or using
560 local_scan() because the Received: header was not generated till after they
561 were called (in order to set the time as the time of reception completion).
562 I have revised the way this works. The header is now generated after the
563 body is received, but before the ACL or local_scan() are called. After they
564 are run, the timestamp in the header is updated.
570 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
571 before starting a queue runner without re-exec. This happened only when
572 deliver_drop_privilege was set or when the Exim user was set to root. The
573 effect of the bug was that timeouts during subsequent deliveries caused
574 crashes instead of being properly handled. The handler is now left at its
575 default (and expected) setting.
577 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
578 message, again when deliver_drop_privilege is set or Exim is run as root.
579 The bug described in (1) was not present in this case, but the tidying up
580 of the other signals was missing. I have made the two cases consistent.
582 3. The ignore_target_hosts setting on a manualroute router was being ignored
583 for hosts that were looked up using the /MX notation.
585 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
588 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
589 operated on the sender address. After changing the $sender_address to <>
590 for the sender address verify, Exim was re-instated it as the original
591 (before rewriting) address, but remembering that it had rewritten it, so it
592 wasn't rewriting it again. This bug also had the effect of breaking the
593 sender address verification caching when the sender address was rewritten.
595 6. The ignore_target_hosts option was being ignored by the ipliteral router.
596 This has been changed so that if the ip literal address matches
597 ignore_target_hosts, the router declines.
599 7. Added expansion conditions match_domain, match_address, and match_local_
600 part (NOT match_host).
602 8. The placeholder for the Received: header didn't have a length field set.
604 9. Added code to Exim itself and to exim_lock to test for a specific race
605 condition that could lead to file corruption when using MBX delivery. The
606 issue is with the lockfile that is created in /tmp. If this file is removed
607 after a process has opened it but before that process has acquired a lock,
608 there is the potential for a second process to recreate the file and also
609 acquire a lock. This could lead to two Exim processes writing to the file
610 at the same time. The added code performs the same test as UW imapd; it
611 checks after acquiring the lock that its file descriptor still refers to
614 10. The buffer for building added header lines was of fixed size, 8192 bytes.
615 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
618 11. Added the smtp_active_hostname option. If used, this will typically be made
619 to depend on the incoming interface address. Because $interface_address is
620 not set up until the daemon has forked a reception process, error responses
621 that can happen earlier (such as "too many connections") no longer contain
624 12. If an expansion in a condition on a "warn" statement fails because a lookup
625 defers, the "warn" statement is abandoned, and the next ACL statement is
626 processed. Previously this caused the whole ACL to be aborted.
628 13. Added the iplsearch lookup type.
630 14. Added ident_timeout as a log selector.
632 15. Added tls_certificate_verified as a log selector.
634 16. Added a global option tls_require_ciphers (compare the smtp transport
635 option of the same name). This controls incoming TLS connections.
637 17. I finally figured out how to make tls_require_ciphers do a similar thing
638 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
639 before starting the TLS session.
641 18. Tabs are now shown as \t in -bP output.
643 19. If the log selector return_path_on_delivery was set, Exim crashed when
644 bouncing a message because it had too many Received: header lines.
646 20. If two routers both had headers_remove settings, and the first one included
647 a superfluous trailing colon, the final name in the first list and the
648 first name in the second list were incorrectly joined into one item (with a
649 colon in the middle).
655 1. Added -C and -D options to the exinext utility, mainly to make it easier
656 to include in the automated testing, but these could be helpful when
657 multiple configurations are in use.
659 2. The exinext utility was not formatting the output nicely when there was
660 an alternate port involved in the retry record key, nor when there was a
661 message id as well (for retries that were specific to a specific message
662 and a specific host). It was also confused by IPv6 addresses, because of
663 the additional colons they contain. I have fixed the IPv4 problem, and
664 patched it up to do a reasonable job for IPv6.
666 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
667 delivery, the log line now contains "pipelined" if PIPELINING was used.
669 4. An SMTP transport process used to panic and die if the bind() call to set
670 an explicit outgoing interface failed. This has been changed; it is now
671 treated in the same way as a connect() failure.
673 5. A reference to $sender_host_name in the part of a conditional expansion
674 that was being skipped was still causing a DNS lookup. This no longer
677 6. The def: expansion condition was not recognizing references to header lines
678 that used bh_ and bheader_.
680 7. Added the _cache feature to named lists.
682 8. The code for checking quota_filecount in the appendfile transport was
683 allowing one more file than it should have been.
685 9. For compatibility with Sendmail, the command line option
693 and sets the incoming protocol and host name (for trusted callers). The
694 host name and its colon can be omitted when only the protocol is to be set.
695 Note the Exim already has two private options, -pd and -ps, that refer to
696 embedded Perl. It is therefore impossible to set a protocol value of "d" or
697 "s", but I don't think that's a major issue.
699 10. A number of refactoring changes to the code, none of which should affect
702 (a) The number of logging options was getting close to filling up the
703 32-bit word that was used as a bit map. I have split them into two classes:
704 those that are passed in the argument to log_write(), and those that are
705 only ever tested independently outside of that function. These are now in
706 separate 32-bit words, so there is plenty of room for expansion again.
707 There is no change in the user interface or the logging behaviour.
709 (b) When building, for example, log lines, the code previously used a
710 macro that called string_cat() twice, in order to add two strings. This is
711 not really sufficiently general. Furthermore, there was one instance where
712 it was actually wrong because one of the argument was used twice, and in
713 one call a function was used. (As it happened, calling the function twice
714 did not affect the overall behaviour.) The macro has been replaced by a
715 function that can join an arbitrary number of extra strings onto a growing
718 (c) The code for expansion conditions now uses a table and a binary chop
719 instead of a serial search (which was left over from when there were very
720 few conditions). Also, it now recognizes conditions like "pam" even when
721 the relevant support is not compiled in: a suitably worded error message is
722 given if an attempt is made to use such a condition.
724 11. Added ${time_interval:xxxxx}.
726 12. A bug was causing one of the ddress fields not to be passed back correctly
727 from remote delivery subprocesses. The field in question was not being
728 subsequently used, so this caused to problems in practice.
730 13. Added new log selectors queue_time and deliver_time.
732 14. Might have fixed a bug in maildirsizefile handling that threw up
733 "unexpected character" debug warnings, and recalculated the data
734 unnecessarily. In any case, I expanded the warning message to give more
737 15. Added the message "Restricted characters in address" to the statements in
738 the default ACL that block characters like @ and % in local parts.
740 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
741 Three changes have been made:
743 (a) There was a serious bug; a negative response to MAIL caused the whole
744 recipient domain to be cached as invalid, thereby blocking all messages
745 to all local parts at the same domain, from all senders. This bug has
746 been fixed. The domain is no longer cached after a negative response to
747 MAIL if the sender used is not empty.
749 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
752 (c) A new callout option, "use_sender" has been added for people who want
753 the modified behaviour.
759 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
762 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
763 indeed breaks things for older releases.
765 3. Added additional logging to the case where there is a problem reading data
766 from a filter that is running in a subprocess using a pipe, in order to
767 try to track down a specific problem.
769 4. Testing facility fudge: when running in the test harness and attempting
770 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
771 getting "No route to host". Convert this to a timeout.
773 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
776 6. Some OS don't have socklen_t but use size_t instead. This affects the
777 fifth argument of getsockopt() amongst other things. This is now
778 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
779 can be set for individual OS. I have set it for SunOS5, OSF1, and
780 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
781 some earlier ones do not.
783 7. Change 4.30/15 was not doing the test caselessly.
785 8. The standard form for an IPv6 address literal was being rejected by address
786 parsing in, for example, MAIL and RCPT commands. An example of this kind of
787 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
788 this, as well as the form without the "IPv6" on the front (but only when
789 address literals are enabled, of course).
791 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
793 10. Exim crashed if a message with an empty sender address specified by -f
794 encountered a router with an errors_to setting. This could be provoked only
799 where an empty string was supplied; "<>" did not hit this bug.
801 11. Installed PCRE release 4.5.
803 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
804 remained set. It is now erased.
806 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
807 times from message ids (which are base 36 rather than the normal 62).
809 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
810 were being counted as actual protocol errors, and logged if the log
811 selector +smtp_protocol_error was set. One cannot be perfect in this test,
812 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
813 and DATA following a set of rejected RCPTs do not count as protocol errors.
814 In other words, Exim assumes they were pipelined, though this may not
815 actually be the case. Of course, in all cases the client gets an
816 appropriate error code.
818 15. If a lookup fails in an ACL condition, a message about the failure may
819 be available; it is used if testing the ACL cannot continue, because most
820 such messages specify what the cause of the deferral is. However, some
821 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
822 that caused an old message to be retained and used if a later statement
823 caused a defer, replacing the real cause of the deferral.
825 16. If an IP address had so many PTR records that the DNS lookup buffer
826 was not large enough to hold them, Exim could crash while trying to process
827 the truncated data. It now detects and logs this case.
829 17. Further to 4.21/58, another change has been made: if (and only if) the
830 first line of a message (the first header line) ends with CRLF, a bare LF
831 in a subsequent header line has a space inserted after it, so as not to
832 terminate the header.
834 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
835 unnecessarily, used atoi() instead of strtol(), and didn't check the
836 termination when getting file sizes from file names by regex.
838 19. Completely re-implemented the support for maildirsize files, in the light
839 of a number of problems with the previous contributed implementation
840 (4.30/29). In particular:
842 . If the quota is zero, the maildirsize file is maintained, but no quota is
845 . If the maildir directory does not exist, it is created before any attempt
846 to write a maildirsize file.
848 . The quota value in the file is just a cache; if the quota is changed in
849 the transport, the new value overrides.
851 . A regular expression is available for excluding directories from the
854 20. The autoreply transport checks the characters in options that define the
855 message's headers; it allows continued headers, but it was checking with
856 isspace() after an embedded newline instead of explicitly looking for a
859 21. If all the "regular" hosts to which an address was routed had passed their
860 expiry times, and had not reached their retry times, the address was
861 bounced, even if fallback hosts were defined. Now Exim should go on to try
864 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
865 equivalent code in the SMTP transport. Some hosts send humungous responses
866 to HELO/EHLO, more than 1024 it seems.
868 23. Refactoring: code in filter.c used (void *) for "any old type" but this
869 gives compiler warnings in some environments. I've now done it "properly",
872 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
873 (because of problems with the built-in one) was declared to return uschar *
874 instead of char *, causing compiler failure.
876 25. Fixed a file descriptor leak when processing alias/forward files.
878 26. Fixed a minor format string issue in dbfn.c.
880 27. Typo in exim.c: ("dmbnz" for "dbmnz").
882 28. If a filter file refered to $h_xxx or $message_headers, and the headers
883 contained RFC 2047 "words", Exim's memory could, under certain conditions,
886 29. When a sender address is verified, it is cached, to save repeating the test
887 when there is more than one recipient in a message. However, when the
888 verification involves a callout, it is possible for different callout
889 options to be set for different recipients. It is too complicated to keep
890 track of this in the cache, so now Exim always runs a verification when a
891 callout is required, relying on the callout cache for the optimization.
892 The overhead is duplication of the address routing, but this should not be
895 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
896 to be verified with callout=postmaster, and the main callout worked but the
897 postmaster check failed, the verification correctly failed. However, if a
898 subsequent RCPT command asked for sender verification *without* the
899 postmaster check, incorrect caching caused this verification also to fail,
902 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
903 it was not caching the DNS options (qualify_single, search_parents) that
904 were used when the lookup failed. A subsequent lookup with different
905 options therefore always gave the same answer, though there were cases
906 where it should not have. (Example: a "domains = !$mx_any" option on a
907 dnslookup router: the "domains" option is always processed without any
908 widening, but the router might have qualify_single set.) Now Exim uses the
909 cached value only when the same options are set.
911 32. Added John Jetmore's "exipick" utility to the distribution.
913 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
914 than a timeout (e.g. a certificate is required, and is not provided), an
915 Exim server now closes the connection immediately. Previously it waited for
916 the client to close - but if the client is SSL, it seems that they each
917 wait for each other, leading to a delay before one of them times out.
919 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
920 maintained 0.8.x compatibility because I don't think many are using it, and
921 it is clearly obsolete.
923 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
926 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
927 client certificate was expired. A simple patch fixes this, though I don't
928 understand the full logic of why the verify callback is called multiple
931 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
932 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
933 which causes problems with some clients (such as the Certicom SSL Plus
934 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
935 disables the coutermeasure allowing Eudora to connect."
937 38. Exim was not checking that a write() to a log file succeeded. This could
938 lead to Bad Things if a log got too big, in particular if it hit a file
939 size limit. Exim now panics and dies if it cannot write to a log file, just
940 as it does if it cannot open a log file.
942 39. Modified OS/Makefile-Linux so that it now contains
944 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
946 The two -D definitions ensure that Exim is compiled with large file
947 support, which makes it possible to handle log files that are bigger than
950 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
951 instance) a domain was checked against a named list that involved a lookup,
952 causing $domain_data to be set, then another domain was checked against the
953 same list, then the first domain was re-checked, the value of $domain_data
954 after the final check could be wrong. In particular, if the second check
955 failed, it could be set empty. This bug probably also applied to
958 41. The strip_trailing_dot option was not being applied to the address given
959 with the -f command-line option.
961 42. The code for reading a message's header from the spool was incrementing
962 $received_count, but never initializing it. This meant that the value was
963 incorrect (doubled) while delivering a message in the same process in which
964 it was received. In the most common configuration of Exim, this never
965 happens - a fresh exec is done - but it can happen when
966 deliver_drop_privilege is set.
968 43. When Exim logs an SMTP synchronization error - client data sent too soon -
969 it now includes up to 150 characters of the unexpected data in the log
972 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
973 and building data strings. The size of both of these buffers was 10 000
974 bytes - far larger than anybody would *ever* want, thought I. Needless to
975 say, somebody hit the limit. I have increased the maximum line length to
976 20 000 and the maximum data length of concatenated lines to 100 000. I have
977 also fixed two bugs, because there was no checking on these buffers. Tsk,
978 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
981 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
982 lsearch lookups. Now it does.
984 46. When parsing a route_list item in a manualroute router, a fixed-length
985 buffer was used for the list of hosts. I made this 1024 bytes long,
986 thinking that nobody would ever have a list of hosts that long. Wrong.
987 Somebody had a whole pile of complicated expansion conditions, and the
988 string was silently truncated, leading to an expansion error. It turns out
989 that it is easier to change to an unlimited length (owing to other changes
990 that have happened since this code was originally written) than to build
991 structure for giving a limitation error. The length of the item that
992 expands into the list of hosts is now unlimited.
994 47. The lsearch lookup could not handle data where the length of text line was
995 more than 4095 characters. Such lines were truncated, leading to shortened
996 data being returned. It should now handle lines of any length.
998 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
999 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
1002 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
1003 replacing strings such as BIN_DIRECTORY in a source file: the replacement
1004 no longer happens in comment lines. A list of replacements is now placed
1005 at the head of all of the source files, except those whose only change is
1006 to replace PERL_COMMAND in the very first #! line.
1008 50. Replaced the slow insertion sort in queue.c, for sorting the list of
1009 messages on the queue, with a bottom-up merge sort, using code contributed
1010 by Michael Haardt. This should make operations like -bp somewhat faster on
1011 large queues. It won't affect queue runners, except when queue_run_in_order
1014 51. Installed eximstats 1.31 in the distribution.
1016 52. Added support for SRV lookups to the dnslookup router.
1018 53. If an ACL referred to $message_body or $message_body_end, the value was not
1019 reset for any messages that followed in the same SMTP session.
1021 54. The store-handling optimization for building very long strings was not
1022 differentiating between the different store pools. I don't think this
1023 actually made any difference in practice, but I've tidied it.
1025 55. While running the routers to verify a sender address, $sender_address
1026 was still set to the sender address. This is wrong, because when routing to
1027 send a bounce to the sender, it would be empty. Therefore, I have changed
1028 it so that, while verifying a sender address, $sender_address is set to <>.
1029 (There is no change to what happens when verifying a recipient address.)
1031 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
1032 target A or AAAA records (if not already returned) without resetting the
1033 qualify_single or search_parents options of the DNS resolver. These are
1034 inappropriate in this case because the targets of MX and SRV records must
1035 be FQDNs. A broken DNS record could cause trouble if it happened to have a
1036 target that, when qualified, matched something in the local domain. These
1037 two options are now turned off when doing these lookups.
1039 57. It seems that at least some releases of Reiserfs (which does not have the
1040 concept of a fixed number of inodes) returns zero and not -1 for the
1041 number of available inodes. This interacted badly with check_spool_inodes,
1042 which assumed that -1 was the "no such thing" setting. What I have done is
1043 to check that the total number of inodes is greater than zero before doing
1044 the test of how many are available.
1046 58. When a "warn" ACL statement has a log_message modifier, the message is
1047 remembered, and not repeated. This is to avoid a lot of repetition when a
1048 message has many recipients that cause the same warning to be written.
1049 Howewer, Exim was preserving the list of already written lines for an
1050 entire SMTP session, which doesn't seem right. The memory is now reset if a
1051 new message is started.
1053 59. The "rewrite" debugging flag was not showing the result of rewriting in the
1054 debugging output unless log_rewrite was also set.
1056 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
1057 of (int)(handle) when we know that handle contains (void *)(-1).
1059 61. The Exim daemon panic-logs an error return when it closes the incoming
1060 connection. However "connection reset by peer" seems to be common, and
1061 isn't really an error worthy of noting specially, so that particular error
1064 62. When Exim is trying to find all the local interfaces, it used to panic and
1065 die if the ioctl to get the interface flags failed. However, it seems that
1066 on at least one OS (Solaris 9) it is possible to have an interface that is
1067 included in the list of interfaces, but for which you get a failure error
1068 for this call. This happens when the interface is not "plumbed" into a
1069 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
1070 failure of the "get flags" call assumes that the interface is down.
1072 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
1073 makes life easier for people who are doing arithmetic on fields extracted
1074 from dates, where you often get leading zeros that should not be
1075 interpreted as octal.
1077 64. Added qualify_domain to the redirect router, to override the global
1080 65. If a pathologically long header line contained very many addresses (the
1081 report of this problem mentioned 10 000) and each of them was rewritten,
1082 Exim could use up a very large amount of memory. (It kept on making new
1083 copies of the header line as it rewrote, and never released the old ones.)
1084 At the expense of a bit more processing, the header rewriting function has
1085 been changed so that it no longer eats memory in this way.
1087 66. The generation of the Received: header has been moved from the time that a
1088 message starts to be received, to the time that it finishes. The timestamp
1089 in the Received: header should now be very close to that of the <= log
1090 line. There are two side-effects of this change:
1092 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
1093 logged header lines no longer include the local Received: line, because
1094 it has not yet been created. The same applies to a copy of the message
1095 that is returned to a non-SMTP sender when a message is rejected.
1097 (b) When a filter file is tested using -bf, no additional Received: header
1098 is added to the test message. After some thought, I decided that this
1101 This change does not affect the value of $received_for. It is still set
1102 after address rewriting, but before local_scan() is called.
1104 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
1106 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
1107 gave an unhelpful panic error message, and a defer error. I have managed to
1108 change this behaviour so that it now rejects any supplied certificate,
1109 which seems right, as the list of acceptable certificates is empty.
1111 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
1112 gave an unhelpful defer error. I have not managed to make this reject any
1113 supplied certificates, but the error message it gives is "no certificate
1114 supplied", which is not helpful.
1116 70. exigrep's output now also includes lines that are not associated with any
1117 message, but which match the given pattern. Implemented by a patch from
1118 Martin Sluka, which also tidied up the Perl a bit.
1120 71. Recipient callout verification, like sender verification, was using <> in
1121 the MAIL FROM command. This isn't really the right thing, since the actual
1122 sender may affect whether the remote host accepts the recipient or not. I
1123 have changed it to use the actual sender in the callout; this means that
1124 the cache record is now keyed on a recipient/sender pair, not just the
1125 recipient address. There doesn't seem to be a real danger of callout loops,
1126 since a callout by the remote host to check the sender would use <>.
1127 [SEE ABOVE: changed after hitting problems.]
1129 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
1130 temporary errors. However, in the case of such a code being given after
1131 the end of a data transmission (i.e. after ".") Exim was failing to write
1132 a retry record for the message. (Yes, there was some broken host that was
1133 actually sending 8xx at this point.)
1135 73. An unknown lookup type in a host list could cause Exim to panic-die when
1136 the list was checked. (An example that provoked this was putting <; in the
1137 middle of a list instead of at the start.) If this happened during a DATA
1138 ACL check, a -D file could be left lying around. This kind of configuration
1139 error no longer causes Exim to die; instead it causes a defer errror. The
1140 incident is still logged to the main and panic logs.
1142 74. Buglet left over from Exim 3 conversion. The message "too many messages
1143 in one connection" was written to the rejectlog but not the mainlog, except
1144 when address rewriting (yes!) was being logged.
1146 75. Added write_rejectlog option.
1148 76. When a system filter was run not as root (that is, when system_filter_user
1149 was set), the values of the $n variables were not being returned to the
1150 main process; thus, they were not subsequently available in the $sn
1153 77. Added +return_path_on_delivery log selector.
1155 78. A connection timeout was being treated differently from recipients deferred
1156 when testing hosts_max_try with a message that was older than the host's
1157 retry timeout. (The host should not be counted, thus allowing all hosts to
1158 be tried at least once before bouncing.) This may have been the cause of an
1159 occasionally reported bug whereby a message would remain on the queue
1160 longer than the retry timeout, but would be bounced if a delivery was
1161 forced. I say "may" because I never totally pinned down the problem;
1162 setting up timeout/retry tests is difficult. See also the next item.
1164 79. The ultimate address timeout was not being applied to errors that involved
1165 a combination of host plus message (for example, a timeout on a MAIL
1166 command). When an address resolved to a number of possible hosts, and they
1167 were not all tried for each delivery (e.g. because of hosts_max_try), a
1168 message could remain on the queue longer than the retry timeout.
1170 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
1173 81. Fixed an obscure SMTP outgoing bug which required at least the following
1174 conditions: (a) there was another message waiting for the same server;
1175 (b) the server returned 5xx to all RCPT commands in the first message so
1176 that the message was not completed; (c) the server dropped the connection
1177 or gave a negative response to the RSET that Exim sends to abort the
1178 transaction. The observed case was a dropped connection after DATA that had
1179 been sent in pipelining mode. That is, the server had advertised PIPELINING
1180 but was not implementing it correctly. The effect of the bug was incorrect
1181 behaviour, such as trying another host, and this could lead to a crash.
1187 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
1188 and daemon.c were passed as pointers to ints; they should have been
1189 pointers to socklen_t variables (which are typically unsigned ints).
1191 2. Some signed/unsigned type warnings in the os.c file for Linux have been
1194 3. Fixed a really odd bug that affected only the testing scheme; patching a
1195 certain fixed string in the binary changed the value of another string that
1196 happened to be identical to the end of the original first string.
1198 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
1199 name", it returns that address as the IP address. On some operating
1200 systems (e.g. Solaris), it also passes back the IP address string as the
1201 "host name". However, on others (e.g. Linux), it passes back an empty
1202 string. Exim wasn't checking for this, and was changing the host name to an
1203 empty string, assuming it had been canonicized.
1205 5. Although rare, it is permitted to have more than one PTR record for a given
1206 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
1207 all the names associated with an address, because they do in Solaris.
1208 However, it seems that they do not in Linux for data that comes from the
1209 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
1210 I found this out when I moved to a new Linux workstation and tried to run
1211 the Exim test suite.
1213 To get round this problem I have changed the code so that it now does its
1214 own call to the DNS to look up PTR records when searching for a host name.
1215 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
1216 addresses that are only in /etc/hosts are still found.
1218 This behaviour is, however, controlled by an option called host_lookup_
1219 order, which defaults to "bydns:byaddr". If people want to use the other
1220 order, or indeed, just use one or the other means of lookup, they can
1221 specify it in this variable.
1223 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
1224 some operating systems, this comes back from gethostbyaddr() as an empty
1225 string, and this is what Exim used to test for. However, it seems that in
1226 other systems, "." is yielded. Exim now tests for this case too.
1228 7. The values of check_spool_space and check_log_space are now held internally
1229 as a number of kilobytes instead of an absolute number of bytes. If a
1230 numbers is specified without 'K' or 'M', it is rounded up to the nearest
1231 kilobyte. This means that much larger values can be stored.
1233 8. Exim monitor: an attempt to get the action menu when not actually pointing
1234 at a message produces an empty menu entitled "No message selected". This
1235 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
1236 no entries in it ("Shell widget menu has zero width and/or height"). So I
1237 have added a single, blank menu entry in this case.
1239 9. Added ${quote_local_part.
1241 10. MIME decoding is now applied to the contents of Subject: header lines when
1244 11. Now that a reference to $sender_host_address automatically causes a reverse
1245 lookup to occur if necessary (4.13/18), there is no need to arrange for a
1246 host lookup before query-style lookups in lists that might use this
1247 variable. This has therefore been abolished, and the "net-" prefix is no
1248 longer necessary for query-style lookups.
1250 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
1251 have been a typo for LFLAGS, so it has been changed.
1253 13. The install script calls Exim with "-C /dev/null" in order to find the
1254 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
1255 to be output. Howeve, since Exim outputs its version number before the
1256 error, it didn't break the script. It just looked ugly. I fixed this by
1257 always allowing "-C /dev/null" if the caller is root.
1259 14. Ignore overlarge ACL variable number when reading spool file - insurance
1260 against a later release with more variables having written the file.
1262 15. The standard form for an IPv6 address literal was being rejected by EHLO.
1263 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1264 this, as well as the form without the "IPv6" on the front.
1266 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
1267 OS/Makefile-Darwin file.
1269 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
1270 only to LDAP libraries that do not have LDAP_OPT_DEREF.
1272 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
1273 variable. However, some OS use ints for inodes. I've added cast to long int
1274 to get rid of the compiler warning.
1276 19. I had forgotten to lock out "/../" in configuration file names when
1277 ALT_CONFIG_PREFIX was set.
1279 20. Routers used for verification do not need to specify transports. However,
1280 if such a router generated a host list, and callout was configured, Exim
1281 crashed, because it could not find a port number from the (non-existent)
1282 transport. It now assumes port 25 in this circumstance.
1284 21. Added the -t option to exigrep.
1286 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
1287 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
1288 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
1289 them is set, LOOKUP_LSEARCH is forced.
1291 23. "exim -bV" now outputs a list of lookups that are included in the binary.
1293 24. Added sender and host information to the "rejected by local_scan()" log
1294 line; previously there was no indication of these.
1296 25. Added .include_if_exists.
1298 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
1299 Linux. It turns out that not all file systems support this. Apparently some
1300 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
1301 do it.) To cope with this, the error EINVAL, which means that sync-ing is
1302 not supported on the file descriptor, is now ignored when Exim is trying to
1303 sync a directory. This applies only to Linux.
1305 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
1307 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
1308 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
1311 29. Added support for maildirsize files from supplied patch (modified a bit).
1313 30. The use of :fail: followed by an empty string could lead Exim to respond to
1314 sender verification failures with (e.g.):
1316 550 Verification failed for <xxx>
1317 550 Sender verify failed
1319 where the first response line was missing the '-' that indicates it is not
1320 the final line of the response.
1322 31. The loop for finding the name of the user that called Exim had a hardwired
1323 limit of 10; it now uses the value of finduser_retries, which is used for
1324 all other user lookups.
1326 32. Added $received_count variable, available in data and not_smtp ACLs, and at
1329 33. Exim was neglecting to zero errno before one call of strtol() when
1330 expanding a string and expecting an integer value. On some systems this
1331 resulted in spurious "integer overflow" errors. Also, it was casting the
1332 result into an int without checking.
1334 34. Testing for a connection timeout using "timeout_connect" in the retry rules
1335 did not work. The code looks as if it has *never* worked, though it appears
1336 to have been documented since at least releast 1.62. I have made it work.
1338 35. The "timeout_DNS" error in retry rules, also documented since at least
1339 1.62, also never worked. As it isn't clear exactly what this means, and
1340 clearly it isn't a major issue, I have abolished the feature by treating it
1341 as "timeout", and writing a warning to the main and panic logs.
1343 36. The display of retry rules for -brt wasn't always showing the error code
1346 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
1347 timeout_connect_A, timeout_connect_MX.
1349 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
1350 to the empty sender.
1352 39. The daemon was not analysing the content of -oX till after it had closed
1353 stderr and disconnected from the controlling terminal. This meant that any
1354 syntax errors were only noted on the panic log, and the return code from
1355 the command was 0. By re-arranging the code a little, I've made the
1356 decoding happen first, so such errors now appear on stderr, and the return
1357 code is 1. However, the actual setting up of the sockets still happens in
1358 the disconnected process, so errors there are still only recorded on the
1361 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
1362 connections (as happens on some IP stacks) was logged at start up time as
1363 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
1364 it from "IPv6 and IPv4", which means that two separate sockets are being
1367 41. The debug output for gethostbyname2() or getipnodebyname() failures now
1368 says whether AF_INET or AF_INET6 was passed as an argument.
1370 42. Exiwhat output was messed up when time zones were included in log
1373 43. Exiwhat now gives more information about the daemon's listening ports,
1374 and whether -tls-on-connect was used.
1376 44. The "port" option of the smtp transport is now expanded.
1378 45. A "message" modifier in a "warn" statement in a non-message ACL was being
1379 silently ignored. Now an error message is written to the main and panic
1382 46. There's a new ACL modifier called "logwrite" which writes to a log file
1383 as soon as it is encountered.
1385 47. Added $local_user_uid and $local_user_gid at routing time.
1387 48. Exim crashed when trying to verify a sender address that was being
1390 49. Exim was recognizing only a space character after ".include". It now also
1391 recognizes a tab character.
1393 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
1394 extracting the relevant information from the specification. The man page no
1395 longer contains scrambled data for the -d option, and I've added a section
1396 at the front about calling Exim under different names.
1398 51. Added "extra_headers" argument to the "mail" command in filter files.
1400 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
1403 53. Installed eximstats 1.29.
1405 54. Added transport_filter_timeout as a generic transport option.
1407 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
1408 Cc: header lines. This was required by RFC 822, but it not required by RFC
1411 56. Exim used to add From:, Date:, and Message-Id: header lines to any
1412 incoming messages that did not have them. Now it does so only if the
1413 message originates locally, that is, if there is no associated remote host
1414 address. When Resent- header lines are present, this applies to the Resent-
1415 lines rather than the non-Resent- lines.
1417 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
1418 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
1420 58. Messages for configuration errors now include the name of the main
1421 configuration file - useful now that there may be more than one file in a
1422 list (.included file names were always shown).
1424 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
1425 for those rare installations that do not start the daemon as root or run it
1426 setuid root. I've cut out the call to initgroups() if the daemon is not
1429 60. The Exim user and group can now be bound into the binary as text strings
1430 that are looked up at the start of Exim's processing.
1432 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
1434 62. Added $mailstore_basename variable.
1436 63. Installed patch to sieve.c from Michael Haardt.
1438 64. When Exim failed to open the panic log after failing to open the main log,
1439 the original message it was trying to log was written to stderr and debug
1440 output, but if they were not available (the usual case in production), it
1441 was lost. Now it is written to syslog before the two lines that record the
1442 failures to open the logs.
1444 65. Users' Exim filters run in subprocesses under the user's uid. It is
1445 possible for a "deliver" command or an alias in a "personal" command to
1446 provoke an address rewrite. If logging of address rewriting is configured,
1447 this fails because the process is not running as root or exim. There may be
1448 a better way of dealing with this, but for the moment (because 4.30 needs
1449 to be released), I have disabled address rewrite logging when running a
1450 filter in a non-root, non-exim process.
1456 1. The buildconfig auxiliary program wasn't quoting the value set for
1457 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
1458 not defined. This bug was masked in 4.22 by the effect that was fixed in
1461 2. Some messages that were rejected after a message id was allocated were
1462 shown as "incomplete" by exigrep. It no longer does this for messages that
1463 are rejected by local_scan() or the DATA or non-SMTP ACLs.
1465 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
1466 have allow_domain_literals set, the ID did not get logged in the <= line.
1467 Domain literals are now always recognized in Message-ID: header lines.
1469 4. The first argument for a ${extract expansion item is the key name or field
1470 number. Leading and trailing spaces in this item were not being ignored,
1471 causing some misleading effects.
1473 5. When deliver_drop_privilege was set, single queue runner processes started
1474 manually (i.e. by the command "exim -q") or by the daemon (which uses the
1475 same command in the process it spins off) were not dropping privilege.
1477 6. When the daemon running as "exim" started a queue runner, it always
1478 re-executed Exim in the spun-off process. This is a waste of effort when
1479 deliver_drop_privilege is set. The new process now just calls the
1480 queue-runner function directly.
1486 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
1489 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
1490 ignored. Though the use of -oP was forcing the writing of a pid file, it
1491 was always written to the default place.
1493 3. If the message "no IP address found for host xxxx" is generated during
1494 incoming verification, it is now followed by identification of the incoming
1495 connection (so you can more easily find what provoked it).
1497 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
1499 5. Added some features to "harden" Exim a bit more against certain attacks:
1501 (a) There is now a build-time option called FIXED_NEVER_USERS that can
1502 be put in Local/Makefile. This is like the never_users runtime option,
1503 but it cannot be overridden. The default setting is "root".
1505 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
1506 prefix string with which any file named in a -C command line option
1509 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
1510 is retained for -C and -D only if the caller of Exim is root. Without
1511 it, the exim user may also use -C and -D and retain privilege.
1513 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
1514 command line option is disabled.
1516 6. Macro names set by the -D option must start with an upper case letter, just
1517 like macro names defined in the configuration file.
1519 7. Added "dereference=" facility to LDAP.
1521 8. Two instances of the typo "uknown" in the source files are fixed.
1523 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
1524 the Configure-Makefile script screwed up while processing it.
1526 10. Incorporated PCRE 4.4.
1528 11. The SMTP synchronization check was not operating right at the start of an
1529 SMTP session. For example, it could not catch a HELO sent before the client
1530 waited for the greeting. There is now a check for outstanding input at the
1531 point when the greeting is written. Because of the duplex, asynchronous
1532 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
1533 way, but not yet received, when the check is performed.
1535 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
1536 on TCP/IP sockets, because this apparently causes some broken clients to
1539 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
1540 unchanged) from the Cygwin maintainer.
1542 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
1543 was supported instead of testing for mbx. Effectively a typo.
1545 15. The spa authenticator server code was not checking that the input it
1546 received was valid base64.
1548 16. The debug output line for the "set" modifier in ACLs was not showing the
1549 name of the variable that was being set.
1551 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
1553 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
1554 lookup on its own. Something else was needed to trigger the lookup. For
1555 example, a match in host_lookup or the need for a host name in a host list.
1556 Now, if $sender_host_name is referenced and the host name has not yet been
1557 looked up, a lookup is performed. If the lookup fails, the variable remains
1558 empty, and $host_lookup_failed is set to "1".
1560 19. Added "eqi" as a case-independent comparison operator.
1562 20. The saslauthd authentication condition could segfault if neither service
1563 nor realm was specified.
1565 21. If an overflowing value such as "2048M" was set for message_size_limit, the
1566 error message that was logged was misleading, and incoming SMTP
1567 connections were dropped. The message is now more accurate, and temporary
1568 errors are given to SMTP connections.
1570 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
1571 (except RSET) with a 421 error, until QUIT is received. However, it was
1572 failing to send a response to QUIT.
1574 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
1575 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
1576 ACL is now run after the helo_try_verify_hosts code.
1578 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
1579 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
1580 case-independent, so other case variants are also recognized). Apparently
1581 some systems use these upper case variants.
1583 25. If more than two messages were waiting for the same host, and a transport
1584 filter was specified for the transport, Exim sent two messages over the
1585 same TCP/IP connection, and then failed with "socket operation on non-
1586 socket" when it tried to send the third.
1588 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
1590 27. The extern definition of crypt16() in expand.c was not being excluded when
1591 the OS had its own crypt16() function.
1593 28. Added bounce_return_body as a new option, and bounce_return_size_limit
1594 as a preferred synonym for return_size_limit, both as an option and as an
1597 29. Added LIBS=-liconv to OS/Makefile-OSF1.
1599 30. Changed the default configuration ACL to relax the local part checking rule
1600 for addresses that are not in any local domains. For these addresses,
1601 slashes and pipe symbols are allowed within local parts, but the sequence
1602 /../ is explicitly forbidden.
1604 31. SPA server authentication was not clearing the challenge buffer before
1607 32. log_message in a "warn" ACL statement was writing to the reject log as
1608 well as to the main log, which contradicts the documentation and doesn't
1609 seem right (because no rejection is happening). So I have stopped it.
1611 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
1612 However, I am unable to do any testing of this.
1614 34. Fixed an infelicity in the appendfile transport. When checking directories
1615 for a mailbox, to see if any needed to be created, it was accidentally
1616 using path names with one or more superfluous leading slashes; tracing
1617 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
1619 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
1620 contents are added to the log line that is written for every discarded
1621 recipient. (Previously a log_message setting was ignored.)
1623 36. The ${quote: operator now quotes the string if it is empty.
1625 37. The install script runs exim in order to find its version number. If for
1626 some reason other than non-existence or emptiness, which it checks, it
1627 could not run './exim', it was installing it with an empty version number,
1628 i.e. as "exim-". This error state is now caught, and the installation is
1631 38. An argument was missing from the function that creates an error message
1632 when Exim fails to connect to the socket for saslauthd authentication.
1633 This could cause Exim to crash, or give a corrupted message.
1635 39. Added isip, isip4, and isip6 to ${if conditions.
1637 40. The ACL variables $acl_xx are now saved with the message, and can be
1638 accessed later in routers, transports, and filters.
1640 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
1641 strings in the file are not string-expanded.
1643 42. If a MAIL command specified a SIZE value that was too large to fit into an
1644 int variable, the check against message_size_limit failed. Such values are
1645 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
1646 day this will have to be increased, but I don't think I want to be around
1647 when emails are that large.
1654 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
1655 iconv() is not standard in FreeBSD.
1657 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
1658 IPv6 enabled. The observed symptom was a segmentation fault on return from
1659 the function os_common_find_running_interfaces() in src/os.c.
1661 3. In the check_special_case() function in daemon.c I had used "errno" as an
1662 argument name, which causes warnings on some systems. This was basically a
1663 typo, since it was named "eno" in the comments!
1665 4. The code that waits for the clock to tick (at a resolution of some fraction
1666 of a second) so as to ensure message-id uniqueness was always waiting for
1667 at least one whole tick, when it could have waited for less. [This is
1668 almost certainly not relevant at current processor speeds, where it is
1669 unlikely to ever wait at all. But we try to future-proof.]
1671 5. The function that sleeps for a time interval that includes fractions of a
1672 second contained a race. It did not block SIGALRM between setting the
1673 timer, and suspending (a couple of lines later). If the interval was short
1674 and the sigsuspend() was delayed until after it had expired, the suspension
1675 never ended. On busy systems this could lead to processes getting stuck for
1678 6. Some uncommon configurations may cause a lookup to happen in a queue runner
1679 process, before it forks any delivery processes. The open lookup caching
1680 mechanism meant that the open file or database connection was passed into
1681 the delivery process. The problem was that delivery processes always tidy
1682 up cached lookup data. This could cause a problem for the next delivery
1683 process started by the queue runner, because the external queue runner
1684 process does not know about the closure. So the next delivery process
1685 still has data in the lookup cache. In the case of a file lookup, there was
1686 no problem because closing a file descriptor in a subprocess doesn't affect
1687 the parent. However, if the lookup was caching a connection to a database,
1688 the connection was closed, and the second delivery process was likely to
1689 see errors such as "PGSQL: query failed: server closed the connection
1690 unexpectedly". The problem has been fixed by closing all cached lookups
1691 in a queue runner before running a delivery process.
1693 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
1694 seem to have the "const" qualifier which it has on other OS. I've
1697 8. Change 4.21/2 was too strict. It is only if there are two authenticators
1698 *of the same type* (client or server) with the same public name that an
1699 error should be diagnosed.
1701 9. When Exim looked up a host name for an IP address, but failed to find the
1702 original IP address when looking up the host name (a safety check), it
1703 output the message "<ip address> does not match any IP for NULL", which was
1704 confusing, to say the least. The bug was that the host name should have
1705 appeared instead of "NULL".
1707 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
1708 user that is built into the binary, and the -C or -D options is used, root
1709 privilege is dropped before the configuration file is read. In addition,
1710 logging is switched to stderr instead of the normal log files. If the
1711 configuration then re-defines the Exim user, the unprivileged environment
1712 is probably not what is expected, so Exim logs a panic warning message (but
1715 However, if deliver_drop_privilege is set, the unprivileged state may well
1716 be exactly what is intended, so the warning has been cut out in that case,
1717 and Exim is allowed to try to write to its normal log files.
1723 1. smtp_return_error_details was not giving details for temporary sender
1724 or receiver verification errors.
1726 2. Diagnose a configuration error if two authenticators have the same public
1729 3. Exim used not to create the message log file for a message until the first
1730 delivery attempt. This could be confusing when incoming messages were held
1731 for policy or load reasons. The message log file is now created at the time
1732 the message is received, and an initial "Received" line is written to it.
1734 4. The automatically generated man page for command line options had a minor
1735 bug that caused no ill effects; however, a more serious problem was that
1736 the procedure for building the man page automatically didn't always
1737 operate. Consequently, release 4.20 contains an out-of-date version. This
1738 shouldn't happen again.
1740 5. When building Exim with embedded Perl support, the script that builds the
1741 Makefile was calling 'perl' to find its compile-time parameters, ignoring
1742 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
1744 6. The freeze_tell option was not being used for messages that were frozen on
1745 arrival, either by an ACL or by local_scan().
1747 7. Added the smtp_incomplete_transaction log selector.
1749 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
1750 was accepting AUTH without a new EHLO.
1752 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
1753 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
1756 10. Logging of TCP/IP connections (when configured) now happens in the main
1757 daemon process instead of the child process, so that the TCP/IP connection
1758 count is more accurate (but it can never be perfect).
1760 11. The use of "drop" in a nested ACL was not being handled correctly in the
1761 outer ACL. Now, if condition failure induced by the nested "drop" causes
1762 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
1763 or "require"), the connection is dropped.
1765 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
1766 that yield "discard" can now be used with an "accept" or a "discard" verb,
1767 but an error is generated for any others (because I can't see a useful way
1768 to define what should happen).
1770 13. When an ACL is read dynamically from a file (or anywhere else), the lines
1771 are now processed in the same way as lines in the Exim configuration file.
1772 In particular, continuation lines are supported.
1774 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
1776 15. Added -ti meaning -t -i.
1778 16. Check for letters, digits, hyphens, and dots in the names of dnslist
1779 domains, and warn by logging if others are found.
1781 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
1782 returned from GIFCONF when Exim is trying to find the list of interfaces on
1783 a host. The code in os.c has been modified to copy each ifreq to an aligned
1784 structure in all cases.
1786 Also, in some cases, the returned ifreq's were being copied to a 'struct
1787 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
1788 means the last couple of bytes of an IPv6 address could be chopped if the
1789 ifreq contained only a normal sockaddr (14 bytes storage).
1791 18. Named domain lists were not supported in the hosts_treat_as_local option.
1792 An entry such as +xxxx was not recognized, and was treated as a literal
1795 19. Ensure that header lines added by a DATA ACL are included in the reject log
1796 if the ACL subsequently rejects the message.
1798 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
1799 MD5 (which is deprecated).
1801 21. When testing a filter file using -bf, Exim was writing a message when it
1802 took the sender from a "From " line in the message, but it was not doing so
1803 when it took $return_path from a Return-Path: header line. It now does.
1805 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
1806 with a valid header line field name (a series of printing characters
1807 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
1809 23. Changed "disc" in the source to "disk" to conform to the documentation and
1810 the book and for uniformity.
1812 24. Ignore Sendmail's -Ooption=value command line item.
1814 25. When execve() failed while trying to run a command in a pipe transport,
1815 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
1816 could be confused with a return value of 69 from the command itself. This
1817 has been changed to 127, the value the shell returns if it is asked to run
1818 a non-existent command. The wording for the related log line suggests a
1819 non-existent command as the problem.
1821 26. If received_header_text expands to an empty string, do not add a Received:
1822 header line to the message. (Well, it adds a token one on the spool, but
1823 marks it "old" so that it doesn't get used or transmitted.)
1825 27. Installed eximstats 1.28 (addition of -nt option).
1827 28. There was no check for failure on the call to getsockname() in the daemon
1828 code. This can fail if there is a shortage of resources on the system, with
1829 ENOMEM, for example. A temporary error is now given on failure.
1831 29. Contrary to the C standard, it seems that in some environments, the
1832 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
1833 program. Exim now does this explicitly; it affects the formatting of
1834 timestamps using strftime().
1836 30. If exiqsumm was given junk data, it threw up some uninitialized variable
1837 complaints. I've now initialized all the variables, to avoid this.
1839 32. Header lines added by a system filter were not being "seen" during
1840 transport-time rewrites.
1842 33. The info_callback() function passed to OpenSSL is set up with type void
1843 (*)(SSL *, int, int), as described somewhere. However, when calling the
1844 function (actually a macro) that sets it up, the type void(*)() is
1845 expected. I've put in a cast to prevent warnings from picky compilers.
1847 34. If a DNS black list lookup found a CNAME record, but there were no A
1848 records associated with the domain it pointed at, Exim crashed.
1850 35. If a DNS black list lookup returned more than one A record, Exim ignored
1851 all but the first. It now scans all returned addresses if a particular IP
1852 value is being sought. In this situation, the contents of the
1853 $dnslist_value variable are a list of all the addresses, separated by a
1856 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
1857 to accept a host name and all its aliases if the forward lookup for any of
1858 them yielded the IP address of the incoming connection. Now it accepts only
1859 those names whose forward lookup yields the correct IP address. Any other
1860 names are discarded. This closes a loophole whereby a rogue DNS
1861 administrator could create reverse DNS records to break through a
1862 wildcarded host restriction in an ACL.
1864 37. If a user filter or a system filter that ran in a subprocess used any of
1865 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
1866 the wrong values were passed to the pipe command ($thisaddress had the
1867 value of $0, $0 had the value of $1, etc). This bug was introduced by
1868 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
1870 38. Improved the line breaking for long SMTP error messages from ACLs.
1871 Previously, if there was no break point between 40 and 75 characters, Exim
1872 left the rest of the message alone. Two changes have been made: (a) I've
1873 reduced the minimum length to 35 characters; (b) if it can't find a break
1874 point between 35 and 75 characters, it looks ahead and uses the first one
1875 that it finds. This may give the occasional overlong line, but at least the
1876 remaining text gets split now.
1878 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
1879 file descriptors might be low, and that setting 1000 would always raise it.
1880 It turns out that in some environments, the limit is already over 1000 and
1881 that lowering it causes trouble. So now Exim takes care not to decrease it.
1883 40. When delivering a message, the value of $return_path is set to $sender_
1884 address at the start of routing (routers may change the value). By an
1885 oversight, this default was not being set up when an address was tested by
1886 -bt or -bv, which affected the outcome if any router or filter referred to
1889 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
1890 log only when "message" or "log_message" are set. However, if one of the
1891 conditions was an address verification, or a call to a nested ACL, the
1892 messages generated by the underlying test were being passed through. This
1893 no longer happens. The underlying message is available in $acl_verify_
1894 message for both "message" and "log_message" expansions, so it can be
1895 passed through if needed.
1897 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
1898 new expansion $bh_ to give the encoded byte string without charset
1899 translation. Translation happens only if iconv() is available; HAVE_ICONV
1900 indicates this at build time. HEADERS_CHARSET gives the charset to
1901 translate to; headers_charset can change it in the configuration, and
1902 "headers charset" can change it in an individual filter file.
1904 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
1905 that creates RFC 2047 encoded "words" labels them as that charset instead
1906 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
1907 expansion operator; (ii) when Exim creates a From: line for a local
1908 message; (iii) when a header line is rewritten to include a "phrase" part.
1910 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
1911 buggy, causing it to skip the first lines of messages whose message ID
1912 ended in 'D'. This would not have bitten before Exim release 4.14, because
1913 message IDs were unlikely to end in 'D' before then. The effect was to have
1914 incorrect size information for certain domains.
1916 45. #include "config.h" was missing at the start of the crypt16.c module. This
1917 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
1920 46. If there was a timeout during a "random" callout check, Exim treated it as
1921 a failure of the random address, and carried on sending RSET and the real
1922 address. If the delay was just some slowness somewhere, the response to the
1923 original RCPT would be taken as a response to RSET and so on, causing
1924 mayhem of various kinds.
1926 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
1927 when I implemented it. It didn't allow for the fact that some option values
1928 may legitimatetly be negative (e.g. size_addition), and it didn't even do
1929 the right test for positive values.
1931 48. Domain names in DNS records are case-independent. Exim always looks them up
1932 in lower case. Some resolvers return domain names in exactly the case they
1933 appear in the zone file, that is, they may contain uppercase letters. Not
1934 all resolvers do this - some return always lower case. Exim was treating a
1935 change of case by a resolver as a change of domain, similar to a widening
1936 of a domain abbreviation. This triggered its re-routing code and so it was
1937 trying to route what was effectively the same domain again. This normally
1938 caused routing to fail (because the router wouldn't handle the domain
1939 twice). Now Exim checks for this case specially, and just changes the
1940 casing of the domain that it ultimately uses when it transmits the message
1943 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
1946 50. If a filter generated a file delivery with a non-absolute name (possible if
1947 no home directory exists for the router), the forbid_file option was not
1950 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
1951 the existing equality matching.
1953 52. Exim was using ints instead of ino_t variables in some places where it was
1954 dealing with inode numbers.
1956 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
1957 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
1958 called TMPDIR, and if it finds it is different, it changes its value.
1960 54. The smtp_printf() function is now made available to local_scan() so
1961 additional output lines can be written before returning. There is also an
1962 smtp_fflush() function to enable the detection of a dropped connection.
1963 The variables smtp_input and smtp_batched_input are exported to
1966 55. Changed the default runtime configuration: the message "Unknown user"
1967 has been removed from the ACL, and instead placed on the localuser router,
1968 using the cannot_route_message feature. This means that any verification
1969 failures that generate their own messages won't get overridden. Similarly,
1970 the "Unrouteable address" message that was in the ACL for unverifiable
1971 relay addresses has also been removed.
1973 56. Added hosts_avoid_esmtp to the smtp transport.
1975 57. The exicyclog script was not checking for the esoteric option
1976 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
1977 will work only if exicyclog is run under the appropriate euid.
1979 58. Following a discussion on the list, the rules by which Exim recognises line
1980 endings on incoming messages have been changed. The -dropcr and drop_cr
1981 options are now no-ops, retained only for backwards compatibility. The
1982 following line terminators are recognized: LF CRLF CR. However, special
1983 processing applies to CR:
1985 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
1986 nor a local message in the state where . is a terminator.
1988 (ii) If a bare CR is encountered in a header line, an extra space is added
1989 after the line terminator so as not to end the header. The reasoning
1990 behind this is that bare CRs in header lines are most likely either
1991 to be mistakes, or people trying to play silly games.
1993 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
1994 was being incorrectly given as 18 bytes larger than it should have been.
1995 This is a VOB (very old bug).
1997 60. This may never have affected anything current, but just in case it has:
1998 When the local host is found other than at the start of a list of hosts,
1999 the local host, those with the same MX, and any that follow, are discarded.
2000 When the list in question was part of a longer list of hosts, the following
2001 hosts (not currently being processed) were also being discarded. This no
2002 longer happens. I'm not sure if this situation could ever has previously
2005 61. Added the "/MX" feature to lists of hosts in the manualroute and query
2008 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
2009 header. This is something that is recommended in a new Internet Draft, and
2010 is something that is documented as being done by Sendmail. There are two
2011 possible values. For messages generated by the autoreply transport, Exim
2014 Auto-Submitted: auto-replied
2016 whereas for all other generated messages (e.g. bounces) it adds
2018 Auto-Submitted: auto-generated
2020 63. The "personal" condition in filters now includes a test for the
2021 Auto-Submitted: header. If it contains the string "auto-" the message it
2022 not considered personal.
2024 64. Added rcpt_include_affixes as a generic transport option.
2026 65. Added queue_only_override (default true).
2028 66. Added the syslog_duplication option.
2030 67. If what should have been the first header line of a message consisted of
2031 a space followed by a colon, Exim was mis-interpreting it as a header line.
2032 It isn't of course - it is syntactically invalid and should therefore be
2033 treated as the start of the message body. The misbehaviour could have
2034 caused a number of strange effects, including loss of data in subsequent
2035 header lines, and spool format errors.
2037 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
2038 client host had authenticated. This control can now be exercised by an ACL
2039 for more flexibility.
2041 69. By default, callouts do not happen when testing with -bh. There is now a
2042 variant, -bhc, which does actually run the callout code, including
2043 consulting and updating the callout cache.
2045 70. Added support for saslauthd authentication, courtesy of Alexander
2048 71. If statvfs() failed on the spool or log directories while checking their
2049 size for availability, Exim confusingly gave the error "space shortage".
2050 Furthermore, in debugging mode it crashed with a floating point exception.
2051 These checks are done if check_{spool,log}_{space,inodes} are set, and when
2052 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
2053 serious problem, Exim now writes to the main and panic logs when this
2054 happens, with details of the failure. It then refuses to accept the
2055 incoming message, giving the message "spool directory problem" or "log
2056 directory problem" with a 421 code for SMTP messages.
2058 72. When Exim is about to re-exec itself, it ensures that the file descriptors
2059 0, 1, and 2 exist, because some OS complain for execs without them (see
2060 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
2061 descriptors. However, the code omitted to check that the open succeeded,
2062 causing mysterious errors if for some reason the permissions on /dev/null
2063 got screwed. Now Exim writes a message to the main and panic logs, and
2064 bombs out if it can't open /dev/null.
2066 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
2067 interact so that it is all more flexible. It is supposed to remain
2068 backwards compatible. Also added extra_local_interfaces.
2070 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
2071 to bomb out with an assertion failure - to the client this appears as a
2072 connection drop. This problem occurs in the part of the code that was taken
2073 from the Samba project. Fortunately, the assertion is in a very simple
2074 function, so I have fixed this by reproducing the function inline in the
2075 one place where it is called, and arranging for authentication to fail
2076 instead of killing the process with assert().
2078 75. The SPA client code was not working when the server requested OEM rather
2079 than Unicode encoding.
2081 76. Added code to make require_files with a specific uid setting more usable in
2082 the case where statting the file as root fails - usually a non-root-mounted
2083 NFS file system. When this happens and the failure is EACCES, Exim now
2084 forks a subprocess and does the per-uid checking as the relevant uid.
2086 77. Added process_log_path.
2088 78. If log_file_path was not explicitly set, a setting of check_log_space or
2089 check_log_inodes was ignored.
2091 79. If a space check for the spool or log partitions fails, the incident is now
2092 logged. Of course, in the latter case the data may get lost...
2094 80. Added the %p formatting code to string_format() so that it can be used to
2095 print addresses in debug_print(). Adjusted all the address printing in the
2096 debugging in store.c to use %p rather than %d.
2098 81. There was a concern that a line of code in smtp_in.c could overflow a
2099 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
2100 initially expressed, the concern was not well-founded, because trailing
2101 spaces are removed early. However, if the trailing spaces were followed by
2102 a NULL, they did not get removed, so the overflow was possible. Two fixes
2105 (a) I re-wrote the offending code in a cleaner fashion.
2106 (b) If an incoming SMTP command contains a NULL character, it is rejected
2109 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
2110 runs initgroups(), so that if the Exim user is in any additional groups,
2111 they will be used during message reception.
2117 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob