1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.39 2004/11/24 16:14:50 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.
73 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
74 handles timeouts, both on the server side and network timeouts. Renamed the
75 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
77 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
78 It is now set to "smtps".
80 22. $host_address is now set to the target address during the checking of
83 23. When checking ignore_target_hosts for an ipliteral router, no host name was
84 being passed; this would have caused $sender_host_name to have been used if
85 matching the list had actually called for a host name (not very likely,
86 since this list is usually IP addresses). A host name is now passed as
89 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
90 code that specifies a non-restarting handler (typically sigaction() in
91 modern systems) in an attempt to fix a rare and obscure crash bug.
93 25. Narrowed the window for a race in the daemon that could cause it to ignore
94 SIGCHLD signals. This is not a major problem, because they are used only to
95 wake it up if nothing else does.
97 26. A malformed maildirsize file could cause Exim to calculate negative values
98 for the mailbox size or file count. Odd effects could occur as a result.
99 The maildirsize information is now recalculated if the size or filecount
102 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
103 support for a long time. Removed HAVE_SYS_VFS_H.
105 28. Installed the latest version of exipick from John Jetmore.
107 29. In an address list, if the pattern was not a regular expression, an empty
108 subject address (from a bounce message) matched only if the pattern was an
109 empty string. Non-empty patterns were not even tested. This was the wrong
110 because it is perfectly reasonable to use an empty address as part of a
111 database query. An empty address is now tested by patterns that are
112 lookups. However, all the other forms of pattern expect the subject to
113 contain a local part and a domain, and therefore, for them, an empty
114 address still always fails if the pattern is not itself empty.
116 30. Exim went into a mad DNS loop when attempting to do a callout where the
117 host was specified on an smtp transport, and looking it up yielded more
120 31. Re-factored the code for checking spool and log partition space into a
121 function that finds that data and another that does the check. The former
122 is then used to implement four new variables: $spool_space, $log_space,
123 $spool_inodes, and $log_inodes.
125 32. The RFC2047 encoding function was originally intended for short strings
126 such as real names; it was not keeping to the 75-character limit for
127 encoded words that the RFC imposes. It now respects the limit, and
128 generates multiple encoded words if necessary. To be on the safe side, I
129 have increased the buffer size for the ${rfc2047: expansion operator from
132 33. It is now permitted to omit both strings after an "if" condition; if the
133 condition is true, the result is "true". As before, when the second string
134 is omitted, a false condition yields an empty string. This makes it less
135 cumbersome to write custom ACL and router conditions.
137 34. Failure to deliver a bounce message always caused it to be frozen, even if
138 there was an errors_to setting on the router. The errors_to setting is now
141 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
142 canonical form (fully expanded) before being placed in
143 $sender_host_address.
145 36. The table in the code that translates DNS record types into text (T_A to
146 "A" for instance) was missing entries for NS and CNAME. It is just possible
147 that this could have caused confusion if both these types were looked up
148 for the same domain, because the text type is used as part of Exim's
149 per-process caching. But the chance of anyone hitting this buglet seems
152 37. The dnsdb lookup has been extended in a number of ways.
154 (1) There is a new type, "zns", which walks up the domain tree until it
155 finds some nameserver records. It should be used with care.
157 (2) There is a new type, "mxh", which is like "mx" except that it returns
158 just the host names, not the priorities.
160 (3) It is now possible to give a list of domains (or IP addresses) to be
163 (4) It is now possible to specify the separator character for use when
164 multiple records are returned.
166 38. The dnslists ACL condition has been extended: it is now possible to supply
167 a list of IP addresses and/or domains to be looked up in a particular DNS
170 39. Added log_selector=+queue_time_overall.
172 40. When running the queue in the test harness, wait just a tad after forking a
173 delivery process, to get repeatability of debugging output.
179 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
180 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
181 Now it returns an integer. A number of calls were still expecting a T/F
182 return. Fortuitously, in all cases, the tests worked in OK situations,
183 which is the norm. However, things would have gone wrong on any write
184 failures on the smtp file descriptor. This function is used when sending
185 messages over SMTP and also when doing verify callouts.
187 2. When Exim is called to do synchronous delivery of a locally submitted
188 message (the -odf or -odi options), it no longer closes stderr before doing
191 3. Implemented the mua_wrapper option.
193 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
195 5. Implemented the functions header_remove(), header_testname(),
196 header_add_at_position(), and receive_remove_recipient(), and exported them
199 6. If an ACL "warn" statement specified the addition of headers, Exim already
200 inserted X-ACL-Warn: at the start if there was no header name. However, it
201 was not making this test for the second and subsequent header lines if
202 there were newlines in the string. This meant that an invalid header could
203 be inserted if Exim was badly configured.
205 7. Allow an ACL "warn" statement to add header lines at the start or after all
206 the Received: headers, as well as at the end.
208 8. Added the rcpt_4xx retry error code.
210 9. Added postmaster_mailfrom=xxx to callout verification option.
212 10. Added mailfrom=xxxx to the callout verification option, for verify=
215 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
216 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
218 12. Inserted some casts to stop certain compilers warning when using pointer
219 differences as field lengths or precisions in printf-type calls (mostly
220 affecting debugging statements).
222 13. Added optional readline() support for -be (dynamically loaded).
224 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
225 same clock tick as a message's arrival, so that its received time was the
226 same as the "first fail" time on the retry record, and that message
227 remained on the queue past the ultimate address timeout, every queue runner
228 would try a delivery (because it was past the ultimate address timeout) but
229 after another failure, the ultimate address timeout, which should have then
230 bounced the address, did not kick in. This was a "< instead of <=" error;
231 in most cases the first failure would have been in the next clock tick
232 after the received time, and all would be well.
234 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
235 being recognized when the domain list was tested by the match_domain
236 condition in an expansion string.
238 16. Added the ${str2b64: operator.
240 17. Exim was always calling setrlimit() to set a large limit for the number of
241 processes, without checking whether the existing limit was already
242 adequate. (It did check for the limit on file descriptors.) Furthermore,
243 errors from getrlimit() and setrlimit() were being ignored. Now they are
244 logged to the main and panic logs, but Exim does carry on, to try to do its
245 job under whatever limits there are.
247 18. Imported PCRE 5.0.
249 19. Trivial typo in log message " temporarily refused connection" (the leading
252 20. If the log selector return_path_on_delivery was set and an address was
253 redirected to /dev/null, the delivery process crashed because it assumed
254 that a return path would always be set for a "successful" delivery. In this
255 case, the whole delivery is bypassed as an optimization, and therefore no
258 21. Internal re-arrangement: the function for sending a challenge and reading
259 a response while authentication was assuming a zero-terminated challenge
260 string. It's now changed to take a pointer and a length, to allow for
261 binary data in such strings.
263 22. Added the cyrus_sasl authenticator (code supplied by MBM).
265 23. Exim was not respecting finduser_retries when seeking the login of the
266 uid under which it was called; it was always trying 10 times. (The default
267 setting of finduser_retries is zero.) Also, it was sleeping after the final
268 failure, which is pointless.
270 24. Implemented tls_on_connect_ports.
272 25. Implemented acl_smtp_predata.
274 26. If the domain in control=submission is set empty, Exim assumes that the
275 authenticated id is a complete email address when it generates From: or
276 Sender: header lines.
278 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
279 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
280 chown and chgrp in /bin and hostname in /usr/bin.
282 28. Exim was keeping the "process log" file open after each use, just as it
283 does for the main log. This opens the possibility of it remaining open for
284 long periods when the USR1 signal hits a daemon. Occasional processlog
285 errors were reported, that could have been caused by this. Anyway, it seems
286 much more sensible not to leave this file open at all, so that is what now
289 29. The long-running daemon process does not normally write to the log once it
290 has entered its main loop, and it closes the log before doing so. This is
291 so that log files can straightforwardly be renamed and moved. However,
292 there are a couple of unusual error situations where the daemon does write
293 log entries, and I had neglected to close the log afterwards.
295 30. The text of an SMTP error response that was received during a remote
296 delivery was being truncated at 512 bytes. This is too short for some of
297 the long messages that one sometimes sees. I've increased the limit to
300 31. It is now possible to make retry rules that apply only when a message has a
301 specific sender, in particular, an empty sender.
303 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
304 it possible to be selective about when SMTP synchronization is enforced.
306 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
308 32. Implemented hosts_connection_nolog.
310 33. Added an ACL for QUIT.
312 34. Setting "delay_warning=" to disable warnings was not working; it gave a
315 35. Added mailbox_size and mailbox_filecount to appendfile.
317 36. Added control = no_multiline_responses to ACLs.
319 37. There was a bug in the logic of the code that waits for the clock to tick
320 in the case where the clock went backwards by a substantial amount such
321 that the microsecond fraction of "now" was more than the microsecond
322 fraction of "then" (but the whole seconds number was less).
324 38. Added support for the libradius Radius client library this is found on
325 FreeBSD (previously only the radiusclient library was supported).
331 1. When certain lookups returned multiple values in the form name=value, the
332 quoting of the values was not always being done properly. Specifically:
333 (a) If the value started with a double quote, but contained no whitespace,
335 (b) If the value contained whitespace other than a space character (i.e.
336 tabs or newlines or carriage returns) it was not quoted.
337 This fix has been applied to the mysql and pgsql lookups by writing a
338 separate quoting function and calling it from the lookup code. The fix
339 should probably also be applied to nisplus, ibase and oracle lookups, but
340 since I cannot test any of those, I have not disturbed their existing code.
342 2. A hit in the callout cache for a specific address caused a log line with no
343 reason for rejecting RCPT. Now it says "Previous (cached) callout
344 verification failure".
346 3. There was an off-by-one bug in the queryprogram router. An over-long
347 return line was truncated at 256 instead of 255 characters, thereby
348 overflowing its buffer with the terminating zero. As well as fixing this, I
349 have increased the buffer size to 1024 (and made a note to document this).
351 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
352 when Exim is waiting for an SMTP response from a remote server, Exim
353 restarts its select() call on the socket, thereby resetting its timeout.
354 This is not a problem when such interrupts are rare. Somebody set up a cron
355 job to run exiwhat every 2 minutes, which is less than the normal select()
356 timeout (5 or 10 minutes). This meant that the select() timeout never
357 kicked in because it was always reset. I have fixed this by comparing the
358 time when an interrupt arrives with the time at the start of the first call
359 to select(). If more time than the timeout has elapsed, the interrupt is
360 treated as a timeout.
362 5. Some internal re-factoring in preparation for the addition of Sieve
363 extensions (by MH). In particular, the "personal" test is moved to a
364 separate function, and given an option for scanning Cc: and Bcc: (which is
365 not set for Exim filters).
367 6. When Exim created an email address using the login of the caller as the
368 local part (e.g. when creating a From: or Sender: header line), it was not
369 quoting the local part when it contained special characters such as @.
371 7. Installed new OpenBSD configuration files.
373 8. Reworded some messages for syntax errors in "and" and "or" conditions to
374 try to make them clearer.
376 9. Callout options, other than the timeout value, were being ignored when
377 verifying sender addresses in header lines. For example, when using
379 verify = header_sender/callout=no_cache
381 the cache was (incorrectly) being used.
383 10. Added a missing instance of ${EXE} to the exim_install script; this affects
384 only the Cygwin environment.
386 11. When return_path_on_delivery was set as a log selector, if different remote
387 addresses in the same message used different return paths and parallel
388 remote delivery occurred, the wrong values would sometimes be logged.
389 (Whenever a remote delivery process finished, the return path value from
390 the most recently started remote delivery process was logged.)
392 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
393 lines when AUTH and/or TLS are in use. This is the "received protocol"
394 field. Exim used to use "asmtp" for authenticated SMTP, without any
395 indication (in the protocol name) for TLS use. Now it follows the RFC and
396 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
397 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
398 names appear in log lines as well as in Received: header lines.
400 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
401 extensions, and comparison tests, and to fix some bugs.
403 14. Changes to the "personal" filter test:
405 (1) The test was buggy in that it was just doing the equivalent of
406 "contains" tests on header lines. For example, if a user's address was
407 anne@some.where, the "personal" test would incorrectly be true for
409 To: susanne@some.where
411 This test is now done by extracting each address from the header in turn,
412 and checking the entire address. Other tests that are part of "personal"
413 are now done using regular expressions (for example, to check local parts
414 of addresses in From: header lines).
416 (2) The list of non-personal local parts in From: addresses has been
417 extended to include "listserv", "majordomo", "*-request", and "owner-*",
418 taken from the Sieve specification recommendations.
420 (3) If the message contains any header line starting with "List-" it is
421 treated as non-personal.
423 (4) The test for "circular" in the Subject: header line has been removed
424 because it now seems ill-conceived.
426 15. Minor typos in src/EDITME comments corrected.
428 16. Installed latest exipick from John Jetmore.
430 17. If headers_add on a router specified a text string that was too long for
431 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
432 of string_sprintf() is now avoided.
434 18. $message_body_size was not set (it was always zero) when running the DATA
435 ACL and the local_scan() function.
437 19. For the "mail" command in an Exim filter, no default was being set for
438 the once_repeat time, causing a random time value to be used if "once" was
439 specified. (If the value happened to be <= 0, no repeat happened.) The
440 default is now 0s, meaning "never repeat". The "vacation" command was OK
441 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
442 (I found it when inspecting the code).
444 20. There is now an overall timeout for performing a callout verification. It
445 defaults to 4 times the callout timeout, which applies to individual SMTP
446 commands during the callout. The overall timeout applies when there is more
447 than one host that can be tried. The timeout is checked before trying the
448 next host. This prevents very long delays if there are a large number of
449 hosts and all are timing out (e.g. when the network connections are timing
450 out). The value of the overall timeout can be changed by specifying an
451 additional sub-option for "callout", called "maxwait". For example:
453 verify = sender/callout=5s,maxwait=20s
455 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
456 to the end before writing, but this should make it even safer).
458 22. Exim was forgetting that it had advertised PIPELINING for the second and
459 subsequent messages on an SMTP connection. It was also not resetting its
460 memory on STARTTLS and an internal HELO.
462 23. When Exim logs an SMTP synchronization error within a session, it now
463 records whether PIPELINING has been advertised or not.
465 24. Added 3 instances of "(long int)" casts to time_t variables that were being
466 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
467 rather than long int.
469 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
471 26. Added the never_mail option to autoreply.
477 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
478 crash if the getsockname() call failed; this can happen if a connection is
479 closed very soon after it is established. The problem was simply in the
480 order in which certain operations were done, causing Exim to try to write
481 to the SMTP stream before it had set up the file descriptor. The bug has
482 been fixed by making things happen in the correct order.
488 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
489 before the connection was closed, thus losing the rejection response.
491 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
492 some early Solaris releases, but causes trouble in current releases where
493 socklen_t is defined.
495 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
498 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
501 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
502 that releasing the top of it at the end releases what was used for sub-
503 expansions (unless the block got too big). However, discard this block if
504 the first thing is a variable or header, so that we can use its block when
505 it is dynamic (useful for very large $message_headers, for example).
507 6. Lookups now cache *every* query, not just the most recent. A new, separate
508 store pool is used for this. It can be recovered when all lookup caches are
509 flushed. Lookups now release memory at the end of their result strings.
510 This has involved some general refactoring of the lookup sources.
512 7. Some code has been added to the store_xxx() functions to reduce the amount
513 of flapping under certain conditions.
515 8. log_incoming_interface used to affect only the <= reception log lines. Now
516 it causes the local interface and port to be added to several more SMTP log
517 lines, for example "SMTP connection from", and rejection lines.
519 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
521 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
523 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
524 could be overwritten at the end of the current message (or the start of a
525 new message if it was set in a HELO ACL). The value is now preserved for
526 the duration of the SMTP connection.
528 12. If a transport had a headers_rewrite setting, and a matching header line
529 contained an unqualified address, that address was qualified, even if it
530 did not match any rewriting rules. The underlying bug was that the values
531 of the flags that permit the existence of unqualified sender and recipient
532 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
533 non-local messages, and by -bnq for local messages) were not being
534 preserved with the message after it was received.
536 13. When Exim was logging an SMTP synchronization error, it could sometimes log
537 "next input=" as part of the text comprising the host identity instead of
538 the correct text. The code was using the same buffer for two different
539 strings. However, depending on which order the printing function evaluated
540 its arguments, the bug did not always show up. Under Linux, for example, my
541 test suite worked just fine.
543 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
544 doesn't work with some older versions of Perl. It has been changed to "my",
545 which in any case is probably the better facility to use.
547 15. A really picky compiler found some instances of statements for creating
548 error messages that either had too many or two few arguments for the format
551 16. The size of the buffer for calls to the DNS resolver has been increased
552 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
553 for addresses that have a lot of PTR records. This alleviates a problem; it
554 does not fully solve it.
556 17. A dnsdb lookup for PTR records that receives more data than will fit in the
557 buffer now truncates the list and logs the incident, which is the same
558 action as happens when Exim is looking up a host name and its aliases.
559 Previously in this situation something unpredictable would happen;
560 sometimes it was "internal error: store_reset failed".
562 18. If a server dropped the connection unexpectedly when an Exim client was
563 using GnuTLS and trying to read a response, the client delivery process
564 crashed while trying to generate an error log message.
566 19. If a "warn" verb in an ACL added multiple headers to a message in a single
569 warn message = H1: something\nH2: something
571 the text was added as a single header line from Exim's point of view
572 though it ended up OK in the delivered message. However, searching for the
573 second and subsequent header lines using $h_h2: did not work. This has been
574 fixed. Similarly, if a system filter added multiple headers in this way,
575 the routers could not see them.
577 20. Expanded the error message when iplsearch is called with an invalid key to
578 suggest using net-iplsearch in a host list.
580 21. When running tests using -bh, any delays imposed by "delay" modifiers in
581 ACLs are no longer actually imposed (and a message to that effect is
584 22. If a "gecos" field in a passwd entry contained escaped characters, in
585 particular, if it contained a \" sequence, Exim got it wrong when building
586 a From: or a Sender: header from that name. A second bug also caused
587 incorrect handling when an unquoted " was present following a character
590 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
591 was not being matched caselessly.
593 24. Arranged for all hyphens in the exim.8 source to be escaped with
596 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
597 properly. Recipient callout cache records were still being keyed to include
598 the sender, even when use_sender was set false. This led to far more
599 callouts that were necessary. The sender is no longer included in the key
600 when use_sender is false.
602 26. Added "control = submission" modifier to ACLs.
604 27. Added the ${base62d: operator to decode base 62 numbers.
606 28. dnsdb lookups can now access SRV records.
608 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
609 the configuration file.
611 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
612 (-v) mode. This makes the output for a verbose queue run more intelligible.
614 31. Added a use_postmaster feature to recipient callouts.
616 32. Added the $body_zerocount variable, containing the number of binary zero
617 bytes in the message body.
619 33. The time of last modification of the "new" subdirectory is now used as the
620 "mailbox time last read" when there is a quota error for a maildir
623 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
625 35. Added +ignore_unknown as a special item in host lists.
627 36. Code for decoding IPv6 addresses in host lists is now included, even if
628 IPv6 support is not being compiled. This fixes a bug in which an IPv6
629 address was recognized as an IP address, but was then not correctly decoded
630 into binary, causing unexpected and incorrect effects when compared with
637 1. Very minor rewording of debugging text in manualroute to say "list of
638 hosts" instead of "hostlist".
640 2. If verify=header_syntax was set, and a header line with an unqualified
641 address (no domain) and a large number of spaces between the end of the
642 name and the colon was received, the reception process suffered a buffer
643 overflow, and (when I tested it) crashed. This was caused by some obsolete
644 code that should have been removed. The fix is to remove it!
646 3. When running in the test harness, delay a bit after writing a bounce
647 message to get a bit more predictability in the log output.
649 4. Added a call to search_tidyup() just before forking a reception process. In
650 theory, someone could use a lookup in the expansion of smtp_accept_max_
651 per_host which, without the tidyup, could leave open a database connection.
653 5. Added the variables $recipient_data and $sender_data which get set from a
654 lookup success in an ACL "recipients" or "senders" condition, or a router
655 "senders" option, similar to $domain_data and $local_part_data.
657 6. Moved the writing of debug_print from before to after the "senders" test
660 7. Change 4.31/66 (moving the time when the Received: is generated) caused
661 problems for message scanning, either using a data ACL, or using
662 local_scan() because the Received: header was not generated till after they
663 were called (in order to set the time as the time of reception completion).
664 I have revised the way this works. The header is now generated after the
665 body is received, but before the ACL or local_scan() are called. After they
666 are run, the timestamp in the header is updated.
672 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
673 before starting a queue runner without re-exec. This happened only when
674 deliver_drop_privilege was set or when the Exim user was set to root. The
675 effect of the bug was that timeouts during subsequent deliveries caused
676 crashes instead of being properly handled. The handler is now left at its
677 default (and expected) setting.
679 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
680 message, again when deliver_drop_privilege is set or Exim is run as root.
681 The bug described in (1) was not present in this case, but the tidying up
682 of the other signals was missing. I have made the two cases consistent.
684 3. The ignore_target_hosts setting on a manualroute router was being ignored
685 for hosts that were looked up using the /MX notation.
687 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
690 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
691 operated on the sender address. After changing the $sender_address to <>
692 for the sender address verify, Exim was re-instated it as the original
693 (before rewriting) address, but remembering that it had rewritten it, so it
694 wasn't rewriting it again. This bug also had the effect of breaking the
695 sender address verification caching when the sender address was rewritten.
697 6. The ignore_target_hosts option was being ignored by the ipliteral router.
698 This has been changed so that if the ip literal address matches
699 ignore_target_hosts, the router declines.
701 7. Added expansion conditions match_domain, match_address, and match_local_
702 part (NOT match_host).
704 8. The placeholder for the Received: header didn't have a length field set.
706 9. Added code to Exim itself and to exim_lock to test for a specific race
707 condition that could lead to file corruption when using MBX delivery. The
708 issue is with the lockfile that is created in /tmp. If this file is removed
709 after a process has opened it but before that process has acquired a lock,
710 there is the potential for a second process to recreate the file and also
711 acquire a lock. This could lead to two Exim processes writing to the file
712 at the same time. The added code performs the same test as UW imapd; it
713 checks after acquiring the lock that its file descriptor still refers to
716 10. The buffer for building added header lines was of fixed size, 8192 bytes.
717 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
720 11. Added the smtp_active_hostname option. If used, this will typically be made
721 to depend on the incoming interface address. Because $interface_address is
722 not set up until the daemon has forked a reception process, error responses
723 that can happen earlier (such as "too many connections") no longer contain
726 12. If an expansion in a condition on a "warn" statement fails because a lookup
727 defers, the "warn" statement is abandoned, and the next ACL statement is
728 processed. Previously this caused the whole ACL to be aborted.
730 13. Added the iplsearch lookup type.
732 14. Added ident_timeout as a log selector.
734 15. Added tls_certificate_verified as a log selector.
736 16. Added a global option tls_require_ciphers (compare the smtp transport
737 option of the same name). This controls incoming TLS connections.
739 17. I finally figured out how to make tls_require_ciphers do a similar thing
740 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
741 before starting the TLS session.
743 18. Tabs are now shown as \t in -bP output.
745 19. If the log selector return_path_on_delivery was set, Exim crashed when
746 bouncing a message because it had too many Received: header lines.
748 20. If two routers both had headers_remove settings, and the first one included
749 a superfluous trailing colon, the final name in the first list and the
750 first name in the second list were incorrectly joined into one item (with a
751 colon in the middle).
757 1. Added -C and -D options to the exinext utility, mainly to make it easier
758 to include in the automated testing, but these could be helpful when
759 multiple configurations are in use.
761 2. The exinext utility was not formatting the output nicely when there was
762 an alternate port involved in the retry record key, nor when there was a
763 message id as well (for retries that were specific to a specific message
764 and a specific host). It was also confused by IPv6 addresses, because of
765 the additional colons they contain. I have fixed the IPv4 problem, and
766 patched it up to do a reasonable job for IPv6.
768 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
769 delivery, the log line now contains "pipelined" if PIPELINING was used.
771 4. An SMTP transport process used to panic and die if the bind() call to set
772 an explicit outgoing interface failed. This has been changed; it is now
773 treated in the same way as a connect() failure.
775 5. A reference to $sender_host_name in the part of a conditional expansion
776 that was being skipped was still causing a DNS lookup. This no longer
779 6. The def: expansion condition was not recognizing references to header lines
780 that used bh_ and bheader_.
782 7. Added the _cache feature to named lists.
784 8. The code for checking quota_filecount in the appendfile transport was
785 allowing one more file than it should have been.
787 9. For compatibility with Sendmail, the command line option
795 and sets the incoming protocol and host name (for trusted callers). The
796 host name and its colon can be omitted when only the protocol is to be set.
797 Note the Exim already has two private options, -pd and -ps, that refer to
798 embedded Perl. It is therefore impossible to set a protocol value of "d" or
799 "s", but I don't think that's a major issue.
801 10. A number of refactoring changes to the code, none of which should affect
804 (a) The number of logging options was getting close to filling up the
805 32-bit word that was used as a bit map. I have split them into two classes:
806 those that are passed in the argument to log_write(), and those that are
807 only ever tested independently outside of that function. These are now in
808 separate 32-bit words, so there is plenty of room for expansion again.
809 There is no change in the user interface or the logging behaviour.
811 (b) When building, for example, log lines, the code previously used a
812 macro that called string_cat() twice, in order to add two strings. This is
813 not really sufficiently general. Furthermore, there was one instance where
814 it was actually wrong because one of the argument was used twice, and in
815 one call a function was used. (As it happened, calling the function twice
816 did not affect the overall behaviour.) The macro has been replaced by a
817 function that can join an arbitrary number of extra strings onto a growing
820 (c) The code for expansion conditions now uses a table and a binary chop
821 instead of a serial search (which was left over from when there were very
822 few conditions). Also, it now recognizes conditions like "pam" even when
823 the relevant support is not compiled in: a suitably worded error message is
824 given if an attempt is made to use such a condition.
826 11. Added ${time_interval:xxxxx}.
828 12. A bug was causing one of the ddress fields not to be passed back correctly
829 from remote delivery subprocesses. The field in question was not being
830 subsequently used, so this caused to problems in practice.
832 13. Added new log selectors queue_time and deliver_time.
834 14. Might have fixed a bug in maildirsizefile handling that threw up
835 "unexpected character" debug warnings, and recalculated the data
836 unnecessarily. In any case, I expanded the warning message to give more
839 15. Added the message "Restricted characters in address" to the statements in
840 the default ACL that block characters like @ and % in local parts.
842 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
843 Three changes have been made:
845 (a) There was a serious bug; a negative response to MAIL caused the whole
846 recipient domain to be cached as invalid, thereby blocking all messages
847 to all local parts at the same domain, from all senders. This bug has
848 been fixed. The domain is no longer cached after a negative response to
849 MAIL if the sender used is not empty.
851 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
854 (c) A new callout option, "use_sender" has been added for people who want
855 the modified behaviour.
861 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
864 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
865 indeed breaks things for older releases.
867 3. Added additional logging to the case where there is a problem reading data
868 from a filter that is running in a subprocess using a pipe, in order to
869 try to track down a specific problem.
871 4. Testing facility fudge: when running in the test harness and attempting
872 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
873 getting "No route to host". Convert this to a timeout.
875 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
878 6. Some OS don't have socklen_t but use size_t instead. This affects the
879 fifth argument of getsockopt() amongst other things. This is now
880 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
881 can be set for individual OS. I have set it for SunOS5, OSF1, and
882 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
883 some earlier ones do not.
885 7. Change 4.30/15 was not doing the test caselessly.
887 8. The standard form for an IPv6 address literal was being rejected by address
888 parsing in, for example, MAIL and RCPT commands. An example of this kind of
889 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
890 this, as well as the form without the "IPv6" on the front (but only when
891 address literals are enabled, of course).
893 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
895 10. Exim crashed if a message with an empty sender address specified by -f
896 encountered a router with an errors_to setting. This could be provoked only
901 where an empty string was supplied; "<>" did not hit this bug.
903 11. Installed PCRE release 4.5.
905 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
906 remained set. It is now erased.
908 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
909 times from message ids (which are base 36 rather than the normal 62).
911 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
912 were being counted as actual protocol errors, and logged if the log
913 selector +smtp_protocol_error was set. One cannot be perfect in this test,
914 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
915 and DATA following a set of rejected RCPTs do not count as protocol errors.
916 In other words, Exim assumes they were pipelined, though this may not
917 actually be the case. Of course, in all cases the client gets an
918 appropriate error code.
920 15. If a lookup fails in an ACL condition, a message about the failure may
921 be available; it is used if testing the ACL cannot continue, because most
922 such messages specify what the cause of the deferral is. However, some
923 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
924 that caused an old message to be retained and used if a later statement
925 caused a defer, replacing the real cause of the deferral.
927 16. If an IP address had so many PTR records that the DNS lookup buffer
928 was not large enough to hold them, Exim could crash while trying to process
929 the truncated data. It now detects and logs this case.
931 17. Further to 4.21/58, another change has been made: if (and only if) the
932 first line of a message (the first header line) ends with CRLF, a bare LF
933 in a subsequent header line has a space inserted after it, so as not to
934 terminate the header.
936 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
937 unnecessarily, used atoi() instead of strtol(), and didn't check the
938 termination when getting file sizes from file names by regex.
940 19. Completely re-implemented the support for maildirsize files, in the light
941 of a number of problems with the previous contributed implementation
942 (4.30/29). In particular:
944 . If the quota is zero, the maildirsize file is maintained, but no quota is
947 . If the maildir directory does not exist, it is created before any attempt
948 to write a maildirsize file.
950 . The quota value in the file is just a cache; if the quota is changed in
951 the transport, the new value overrides.
953 . A regular expression is available for excluding directories from the
956 20. The autoreply transport checks the characters in options that define the
957 message's headers; it allows continued headers, but it was checking with
958 isspace() after an embedded newline instead of explicitly looking for a
961 21. If all the "regular" hosts to which an address was routed had passed their
962 expiry times, and had not reached their retry times, the address was
963 bounced, even if fallback hosts were defined. Now Exim should go on to try
966 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
967 equivalent code in the SMTP transport. Some hosts send humungous responses
968 to HELO/EHLO, more than 1024 it seems.
970 23. Refactoring: code in filter.c used (void *) for "any old type" but this
971 gives compiler warnings in some environments. I've now done it "properly",
974 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
975 (because of problems with the built-in one) was declared to return uschar *
976 instead of char *, causing compiler failure.
978 25. Fixed a file descriptor leak when processing alias/forward files.
980 26. Fixed a minor format string issue in dbfn.c.
982 27. Typo in exim.c: ("dmbnz" for "dbmnz").
984 28. If a filter file refered to $h_xxx or $message_headers, and the headers
985 contained RFC 2047 "words", Exim's memory could, under certain conditions,
988 29. When a sender address is verified, it is cached, to save repeating the test
989 when there is more than one recipient in a message. However, when the
990 verification involves a callout, it is possible for different callout
991 options to be set for different recipients. It is too complicated to keep
992 track of this in the cache, so now Exim always runs a verification when a
993 callout is required, relying on the callout cache for the optimization.
994 The overhead is duplication of the address routing, but this should not be
997 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
998 to be verified with callout=postmaster, and the main callout worked but the
999 postmaster check failed, the verification correctly failed. However, if a
1000 subsequent RCPT command asked for sender verification *without* the
1001 postmaster check, incorrect caching caused this verification also to fail,
1004 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1005 it was not caching the DNS options (qualify_single, search_parents) that
1006 were used when the lookup failed. A subsequent lookup with different
1007 options therefore always gave the same answer, though there were cases
1008 where it should not have. (Example: a "domains = !$mx_any" option on a
1009 dnslookup router: the "domains" option is always processed without any
1010 widening, but the router might have qualify_single set.) Now Exim uses the
1011 cached value only when the same options are set.
1013 32. Added John Jetmore's "exipick" utility to the distribution.
1015 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
1016 than a timeout (e.g. a certificate is required, and is not provided), an
1017 Exim server now closes the connection immediately. Previously it waited for
1018 the client to close - but if the client is SSL, it seems that they each
1019 wait for each other, leading to a delay before one of them times out.
1021 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1022 maintained 0.8.x compatibility because I don't think many are using it, and
1023 it is clearly obsolete.
1025 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1028 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1029 client certificate was expired. A simple patch fixes this, though I don't
1030 understand the full logic of why the verify callback is called multiple
1033 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1034 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1035 which causes problems with some clients (such as the Certicom SSL Plus
1036 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1037 disables the coutermeasure allowing Eudora to connect."
1039 38. Exim was not checking that a write() to a log file succeeded. This could
1040 lead to Bad Things if a log got too big, in particular if it hit a file
1041 size limit. Exim now panics and dies if it cannot write to a log file, just
1042 as it does if it cannot open a log file.
1044 39. Modified OS/Makefile-Linux so that it now contains
1046 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
1048 The two -D definitions ensure that Exim is compiled with large file
1049 support, which makes it possible to handle log files that are bigger than
1052 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
1053 instance) a domain was checked against a named list that involved a lookup,
1054 causing $domain_data to be set, then another domain was checked against the
1055 same list, then the first domain was re-checked, the value of $domain_data
1056 after the final check could be wrong. In particular, if the second check
1057 failed, it could be set empty. This bug probably also applied to
1060 41. The strip_trailing_dot option was not being applied to the address given
1061 with the -f command-line option.
1063 42. The code for reading a message's header from the spool was incrementing
1064 $received_count, but never initializing it. This meant that the value was
1065 incorrect (doubled) while delivering a message in the same process in which
1066 it was received. In the most common configuration of Exim, this never
1067 happens - a fresh exec is done - but it can happen when
1068 deliver_drop_privilege is set.
1070 43. When Exim logs an SMTP synchronization error - client data sent too soon -
1071 it now includes up to 150 characters of the unexpected data in the log
1074 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
1075 and building data strings. The size of both of these buffers was 10 000
1076 bytes - far larger than anybody would *ever* want, thought I. Needless to
1077 say, somebody hit the limit. I have increased the maximum line length to
1078 20 000 and the maximum data length of concatenated lines to 100 000. I have
1079 also fixed two bugs, because there was no checking on these buffers. Tsk,
1080 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
1081 buffer is too small.
1083 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
1084 lsearch lookups. Now it does.
1086 46. When parsing a route_list item in a manualroute router, a fixed-length
1087 buffer was used for the list of hosts. I made this 1024 bytes long,
1088 thinking that nobody would ever have a list of hosts that long. Wrong.
1089 Somebody had a whole pile of complicated expansion conditions, and the
1090 string was silently truncated, leading to an expansion error. It turns out
1091 that it is easier to change to an unlimited length (owing to other changes
1092 that have happened since this code was originally written) than to build
1093 structure for giving a limitation error. The length of the item that
1094 expands into the list of hosts is now unlimited.
1096 47. The lsearch lookup could not handle data where the length of text line was
1097 more than 4095 characters. Such lines were truncated, leading to shortened
1098 data being returned. It should now handle lines of any length.
1100 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
1101 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
1104 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
1105 replacing strings such as BIN_DIRECTORY in a source file: the replacement
1106 no longer happens in comment lines. A list of replacements is now placed
1107 at the head of all of the source files, except those whose only change is
1108 to replace PERL_COMMAND in the very first #! line.
1110 50. Replaced the slow insertion sort in queue.c, for sorting the list of
1111 messages on the queue, with a bottom-up merge sort, using code contributed
1112 by Michael Haardt. This should make operations like -bp somewhat faster on
1113 large queues. It won't affect queue runners, except when queue_run_in_order
1116 51. Installed eximstats 1.31 in the distribution.
1118 52. Added support for SRV lookups to the dnslookup router.
1120 53. If an ACL referred to $message_body or $message_body_end, the value was not
1121 reset for any messages that followed in the same SMTP session.
1123 54. The store-handling optimization for building very long strings was not
1124 differentiating between the different store pools. I don't think this
1125 actually made any difference in practice, but I've tidied it.
1127 55. While running the routers to verify a sender address, $sender_address
1128 was still set to the sender address. This is wrong, because when routing to
1129 send a bounce to the sender, it would be empty. Therefore, I have changed
1130 it so that, while verifying a sender address, $sender_address is set to <>.
1131 (There is no change to what happens when verifying a recipient address.)
1133 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
1134 target A or AAAA records (if not already returned) without resetting the
1135 qualify_single or search_parents options of the DNS resolver. These are
1136 inappropriate in this case because the targets of MX and SRV records must
1137 be FQDNs. A broken DNS record could cause trouble if it happened to have a
1138 target that, when qualified, matched something in the local domain. These
1139 two options are now turned off when doing these lookups.
1141 57. It seems that at least some releases of Reiserfs (which does not have the
1142 concept of a fixed number of inodes) returns zero and not -1 for the
1143 number of available inodes. This interacted badly with check_spool_inodes,
1144 which assumed that -1 was the "no such thing" setting. What I have done is
1145 to check that the total number of inodes is greater than zero before doing
1146 the test of how many are available.
1148 58. When a "warn" ACL statement has a log_message modifier, the message is
1149 remembered, and not repeated. This is to avoid a lot of repetition when a
1150 message has many recipients that cause the same warning to be written.
1151 Howewer, Exim was preserving the list of already written lines for an
1152 entire SMTP session, which doesn't seem right. The memory is now reset if a
1153 new message is started.
1155 59. The "rewrite" debugging flag was not showing the result of rewriting in the
1156 debugging output unless log_rewrite was also set.
1158 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
1159 of (int)(handle) when we know that handle contains (void *)(-1).
1161 61. The Exim daemon panic-logs an error return when it closes the incoming
1162 connection. However "connection reset by peer" seems to be common, and
1163 isn't really an error worthy of noting specially, so that particular error
1166 62. When Exim is trying to find all the local interfaces, it used to panic and
1167 die if the ioctl to get the interface flags failed. However, it seems that
1168 on at least one OS (Solaris 9) it is possible to have an interface that is
1169 included in the list of interfaces, but for which you get a failure error
1170 for this call. This happens when the interface is not "plumbed" into a
1171 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
1172 failure of the "get flags" call assumes that the interface is down.
1174 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
1175 makes life easier for people who are doing arithmetic on fields extracted
1176 from dates, where you often get leading zeros that should not be
1177 interpreted as octal.
1179 64. Added qualify_domain to the redirect router, to override the global
1182 65. If a pathologically long header line contained very many addresses (the
1183 report of this problem mentioned 10 000) and each of them was rewritten,
1184 Exim could use up a very large amount of memory. (It kept on making new
1185 copies of the header line as it rewrote, and never released the old ones.)
1186 At the expense of a bit more processing, the header rewriting function has
1187 been changed so that it no longer eats memory in this way.
1189 66. The generation of the Received: header has been moved from the time that a
1190 message starts to be received, to the time that it finishes. The timestamp
1191 in the Received: header should now be very close to that of the <= log
1192 line. There are two side-effects of this change:
1194 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
1195 logged header lines no longer include the local Received: line, because
1196 it has not yet been created. The same applies to a copy of the message
1197 that is returned to a non-SMTP sender when a message is rejected.
1199 (b) When a filter file is tested using -bf, no additional Received: header
1200 is added to the test message. After some thought, I decided that this
1203 This change does not affect the value of $received_for. It is still set
1204 after address rewriting, but before local_scan() is called.
1206 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
1208 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
1209 gave an unhelpful panic error message, and a defer error. I have managed to
1210 change this behaviour so that it now rejects any supplied certificate,
1211 which seems right, as the list of acceptable certificates is empty.
1213 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
1214 gave an unhelpful defer error. I have not managed to make this reject any
1215 supplied certificates, but the error message it gives is "no certificate
1216 supplied", which is not helpful.
1218 70. exigrep's output now also includes lines that are not associated with any
1219 message, but which match the given pattern. Implemented by a patch from
1220 Martin Sluka, which also tidied up the Perl a bit.
1222 71. Recipient callout verification, like sender verification, was using <> in
1223 the MAIL FROM command. This isn't really the right thing, since the actual
1224 sender may affect whether the remote host accepts the recipient or not. I
1225 have changed it to use the actual sender in the callout; this means that
1226 the cache record is now keyed on a recipient/sender pair, not just the
1227 recipient address. There doesn't seem to be a real danger of callout loops,
1228 since a callout by the remote host to check the sender would use <>.
1229 [SEE ABOVE: changed after hitting problems.]
1231 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
1232 temporary errors. However, in the case of such a code being given after
1233 the end of a data transmission (i.e. after ".") Exim was failing to write
1234 a retry record for the message. (Yes, there was some broken host that was
1235 actually sending 8xx at this point.)
1237 73. An unknown lookup type in a host list could cause Exim to panic-die when
1238 the list was checked. (An example that provoked this was putting <; in the
1239 middle of a list instead of at the start.) If this happened during a DATA
1240 ACL check, a -D file could be left lying around. This kind of configuration
1241 error no longer causes Exim to die; instead it causes a defer errror. The
1242 incident is still logged to the main and panic logs.
1244 74. Buglet left over from Exim 3 conversion. The message "too many messages
1245 in one connection" was written to the rejectlog but not the mainlog, except
1246 when address rewriting (yes!) was being logged.
1248 75. Added write_rejectlog option.
1250 76. When a system filter was run not as root (that is, when system_filter_user
1251 was set), the values of the $n variables were not being returned to the
1252 main process; thus, they were not subsequently available in the $sn
1255 77. Added +return_path_on_delivery log selector.
1257 78. A connection timeout was being treated differently from recipients deferred
1258 when testing hosts_max_try with a message that was older than the host's
1259 retry timeout. (The host should not be counted, thus allowing all hosts to
1260 be tried at least once before bouncing.) This may have been the cause of an
1261 occasionally reported bug whereby a message would remain on the queue
1262 longer than the retry timeout, but would be bounced if a delivery was
1263 forced. I say "may" because I never totally pinned down the problem;
1264 setting up timeout/retry tests is difficult. See also the next item.
1266 79. The ultimate address timeout was not being applied to errors that involved
1267 a combination of host plus message (for example, a timeout on a MAIL
1268 command). When an address resolved to a number of possible hosts, and they
1269 were not all tried for each delivery (e.g. because of hosts_max_try), a
1270 message could remain on the queue longer than the retry timeout.
1272 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
1275 81. Fixed an obscure SMTP outgoing bug which required at least the following
1276 conditions: (a) there was another message waiting for the same server;
1277 (b) the server returned 5xx to all RCPT commands in the first message so
1278 that the message was not completed; (c) the server dropped the connection
1279 or gave a negative response to the RSET that Exim sends to abort the
1280 transaction. The observed case was a dropped connection after DATA that had
1281 been sent in pipelining mode. That is, the server had advertised PIPELINING
1282 but was not implementing it correctly. The effect of the bug was incorrect
1283 behaviour, such as trying another host, and this could lead to a crash.
1289 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
1290 and daemon.c were passed as pointers to ints; they should have been
1291 pointers to socklen_t variables (which are typically unsigned ints).
1293 2. Some signed/unsigned type warnings in the os.c file for Linux have been
1296 3. Fixed a really odd bug that affected only the testing scheme; patching a
1297 certain fixed string in the binary changed the value of another string that
1298 happened to be identical to the end of the original first string.
1300 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
1301 name", it returns that address as the IP address. On some operating
1302 systems (e.g. Solaris), it also passes back the IP address string as the
1303 "host name". However, on others (e.g. Linux), it passes back an empty
1304 string. Exim wasn't checking for this, and was changing the host name to an
1305 empty string, assuming it had been canonicized.
1307 5. Although rare, it is permitted to have more than one PTR record for a given
1308 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
1309 all the names associated with an address, because they do in Solaris.
1310 However, it seems that they do not in Linux for data that comes from the
1311 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
1312 I found this out when I moved to a new Linux workstation and tried to run
1313 the Exim test suite.
1315 To get round this problem I have changed the code so that it now does its
1316 own call to the DNS to look up PTR records when searching for a host name.
1317 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
1318 addresses that are only in /etc/hosts are still found.
1320 This behaviour is, however, controlled by an option called host_lookup_
1321 order, which defaults to "bydns:byaddr". If people want to use the other
1322 order, or indeed, just use one or the other means of lookup, they can
1323 specify it in this variable.
1325 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
1326 some operating systems, this comes back from gethostbyaddr() as an empty
1327 string, and this is what Exim used to test for. However, it seems that in
1328 other systems, "." is yielded. Exim now tests for this case too.
1330 7. The values of check_spool_space and check_log_space are now held internally
1331 as a number of kilobytes instead of an absolute number of bytes. If a
1332 numbers is specified without 'K' or 'M', it is rounded up to the nearest
1333 kilobyte. This means that much larger values can be stored.
1335 8. Exim monitor: an attempt to get the action menu when not actually pointing
1336 at a message produces an empty menu entitled "No message selected". This
1337 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
1338 no entries in it ("Shell widget menu has zero width and/or height"). So I
1339 have added a single, blank menu entry in this case.
1341 9. Added ${quote_local_part.
1343 10. MIME decoding is now applied to the contents of Subject: header lines when
1346 11. Now that a reference to $sender_host_address automatically causes a reverse
1347 lookup to occur if necessary (4.13/18), there is no need to arrange for a
1348 host lookup before query-style lookups in lists that might use this
1349 variable. This has therefore been abolished, and the "net-" prefix is no
1350 longer necessary for query-style lookups.
1352 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
1353 have been a typo for LFLAGS, so it has been changed.
1355 13. The install script calls Exim with "-C /dev/null" in order to find the
1356 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
1357 to be output. Howeve, since Exim outputs its version number before the
1358 error, it didn't break the script. It just looked ugly. I fixed this by
1359 always allowing "-C /dev/null" if the caller is root.
1361 14. Ignore overlarge ACL variable number when reading spool file - insurance
1362 against a later release with more variables having written the file.
1364 15. The standard form for an IPv6 address literal was being rejected by EHLO.
1365 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1366 this, as well as the form without the "IPv6" on the front.
1368 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
1369 OS/Makefile-Darwin file.
1371 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
1372 only to LDAP libraries that do not have LDAP_OPT_DEREF.
1374 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
1375 variable. However, some OS use ints for inodes. I've added cast to long int
1376 to get rid of the compiler warning.
1378 19. I had forgotten to lock out "/../" in configuration file names when
1379 ALT_CONFIG_PREFIX was set.
1381 20. Routers used for verification do not need to specify transports. However,
1382 if such a router generated a host list, and callout was configured, Exim
1383 crashed, because it could not find a port number from the (non-existent)
1384 transport. It now assumes port 25 in this circumstance.
1386 21. Added the -t option to exigrep.
1388 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
1389 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
1390 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
1391 them is set, LOOKUP_LSEARCH is forced.
1393 23. "exim -bV" now outputs a list of lookups that are included in the binary.
1395 24. Added sender and host information to the "rejected by local_scan()" log
1396 line; previously there was no indication of these.
1398 25. Added .include_if_exists.
1400 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
1401 Linux. It turns out that not all file systems support this. Apparently some
1402 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
1403 do it.) To cope with this, the error EINVAL, which means that sync-ing is
1404 not supported on the file descriptor, is now ignored when Exim is trying to
1405 sync a directory. This applies only to Linux.
1407 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
1409 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
1410 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
1413 29. Added support for maildirsize files from supplied patch (modified a bit).
1415 30. The use of :fail: followed by an empty string could lead Exim to respond to
1416 sender verification failures with (e.g.):
1418 550 Verification failed for <xxx>
1419 550 Sender verify failed
1421 where the first response line was missing the '-' that indicates it is not
1422 the final line of the response.
1424 31. The loop for finding the name of the user that called Exim had a hardwired
1425 limit of 10; it now uses the value of finduser_retries, which is used for
1426 all other user lookups.
1428 32. Added $received_count variable, available in data and not_smtp ACLs, and at
1431 33. Exim was neglecting to zero errno before one call of strtol() when
1432 expanding a string and expecting an integer value. On some systems this
1433 resulted in spurious "integer overflow" errors. Also, it was casting the
1434 result into an int without checking.
1436 34. Testing for a connection timeout using "timeout_connect" in the retry rules
1437 did not work. The code looks as if it has *never* worked, though it appears
1438 to have been documented since at least releast 1.62. I have made it work.
1440 35. The "timeout_DNS" error in retry rules, also documented since at least
1441 1.62, also never worked. As it isn't clear exactly what this means, and
1442 clearly it isn't a major issue, I have abolished the feature by treating it
1443 as "timeout", and writing a warning to the main and panic logs.
1445 36. The display of retry rules for -brt wasn't always showing the error code
1448 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
1449 timeout_connect_A, timeout_connect_MX.
1451 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
1452 to the empty sender.
1454 39. The daemon was not analysing the content of -oX till after it had closed
1455 stderr and disconnected from the controlling terminal. This meant that any
1456 syntax errors were only noted on the panic log, and the return code from
1457 the command was 0. By re-arranging the code a little, I've made the
1458 decoding happen first, so such errors now appear on stderr, and the return
1459 code is 1. However, the actual setting up of the sockets still happens in
1460 the disconnected process, so errors there are still only recorded on the
1463 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
1464 connections (as happens on some IP stacks) was logged at start up time as
1465 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
1466 it from "IPv6 and IPv4", which means that two separate sockets are being
1469 41. The debug output for gethostbyname2() or getipnodebyname() failures now
1470 says whether AF_INET or AF_INET6 was passed as an argument.
1472 42. Exiwhat output was messed up when time zones were included in log
1475 43. Exiwhat now gives more information about the daemon's listening ports,
1476 and whether -tls-on-connect was used.
1478 44. The "port" option of the smtp transport is now expanded.
1480 45. A "message" modifier in a "warn" statement in a non-message ACL was being
1481 silently ignored. Now an error message is written to the main and panic
1484 46. There's a new ACL modifier called "logwrite" which writes to a log file
1485 as soon as it is encountered.
1487 47. Added $local_user_uid and $local_user_gid at routing time.
1489 48. Exim crashed when trying to verify a sender address that was being
1492 49. Exim was recognizing only a space character after ".include". It now also
1493 recognizes a tab character.
1495 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
1496 extracting the relevant information from the specification. The man page no
1497 longer contains scrambled data for the -d option, and I've added a section
1498 at the front about calling Exim under different names.
1500 51. Added "extra_headers" argument to the "mail" command in filter files.
1502 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
1505 53. Installed eximstats 1.29.
1507 54. Added transport_filter_timeout as a generic transport option.
1509 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
1510 Cc: header lines. This was required by RFC 822, but it not required by RFC
1513 56. Exim used to add From:, Date:, and Message-Id: header lines to any
1514 incoming messages that did not have them. Now it does so only if the
1515 message originates locally, that is, if there is no associated remote host
1516 address. When Resent- header lines are present, this applies to the Resent-
1517 lines rather than the non-Resent- lines.
1519 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
1520 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
1522 58. Messages for configuration errors now include the name of the main
1523 configuration file - useful now that there may be more than one file in a
1524 list (.included file names were always shown).
1526 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
1527 for those rare installations that do not start the daemon as root or run it
1528 setuid root. I've cut out the call to initgroups() if the daemon is not
1531 60. The Exim user and group can now be bound into the binary as text strings
1532 that are looked up at the start of Exim's processing.
1534 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
1536 62. Added $mailstore_basename variable.
1538 63. Installed patch to sieve.c from Michael Haardt.
1540 64. When Exim failed to open the panic log after failing to open the main log,
1541 the original message it was trying to log was written to stderr and debug
1542 output, but if they were not available (the usual case in production), it
1543 was lost. Now it is written to syslog before the two lines that record the
1544 failures to open the logs.
1546 65. Users' Exim filters run in subprocesses under the user's uid. It is
1547 possible for a "deliver" command or an alias in a "personal" command to
1548 provoke an address rewrite. If logging of address rewriting is configured,
1549 this fails because the process is not running as root or exim. There may be
1550 a better way of dealing with this, but for the moment (because 4.30 needs
1551 to be released), I have disabled address rewrite logging when running a
1552 filter in a non-root, non-exim process.
1558 1. The buildconfig auxiliary program wasn't quoting the value set for
1559 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
1560 not defined. This bug was masked in 4.22 by the effect that was fixed in
1563 2. Some messages that were rejected after a message id was allocated were
1564 shown as "incomplete" by exigrep. It no longer does this for messages that
1565 are rejected by local_scan() or the DATA or non-SMTP ACLs.
1567 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
1568 have allow_domain_literals set, the ID did not get logged in the <= line.
1569 Domain literals are now always recognized in Message-ID: header lines.
1571 4. The first argument for a ${extract expansion item is the key name or field
1572 number. Leading and trailing spaces in this item were not being ignored,
1573 causing some misleading effects.
1575 5. When deliver_drop_privilege was set, single queue runner processes started
1576 manually (i.e. by the command "exim -q") or by the daemon (which uses the
1577 same command in the process it spins off) were not dropping privilege.
1579 6. When the daemon running as "exim" started a queue runner, it always
1580 re-executed Exim in the spun-off process. This is a waste of effort when
1581 deliver_drop_privilege is set. The new process now just calls the
1582 queue-runner function directly.
1588 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
1591 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
1592 ignored. Though the use of -oP was forcing the writing of a pid file, it
1593 was always written to the default place.
1595 3. If the message "no IP address found for host xxxx" is generated during
1596 incoming verification, it is now followed by identification of the incoming
1597 connection (so you can more easily find what provoked it).
1599 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
1601 5. Added some features to "harden" Exim a bit more against certain attacks:
1603 (a) There is now a build-time option called FIXED_NEVER_USERS that can
1604 be put in Local/Makefile. This is like the never_users runtime option,
1605 but it cannot be overridden. The default setting is "root".
1607 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
1608 prefix string with which any file named in a -C command line option
1611 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
1612 is retained for -C and -D only if the caller of Exim is root. Without
1613 it, the exim user may also use -C and -D and retain privilege.
1615 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
1616 command line option is disabled.
1618 6. Macro names set by the -D option must start with an upper case letter, just
1619 like macro names defined in the configuration file.
1621 7. Added "dereference=" facility to LDAP.
1623 8. Two instances of the typo "uknown" in the source files are fixed.
1625 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
1626 the Configure-Makefile script screwed up while processing it.
1628 10. Incorporated PCRE 4.4.
1630 11. The SMTP synchronization check was not operating right at the start of an
1631 SMTP session. For example, it could not catch a HELO sent before the client
1632 waited for the greeting. There is now a check for outstanding input at the
1633 point when the greeting is written. Because of the duplex, asynchronous
1634 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
1635 way, but not yet received, when the check is performed.
1637 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
1638 on TCP/IP sockets, because this apparently causes some broken clients to
1641 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
1642 unchanged) from the Cygwin maintainer.
1644 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
1645 was supported instead of testing for mbx. Effectively a typo.
1647 15. The spa authenticator server code was not checking that the input it
1648 received was valid base64.
1650 16. The debug output line for the "set" modifier in ACLs was not showing the
1651 name of the variable that was being set.
1653 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
1655 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
1656 lookup on its own. Something else was needed to trigger the lookup. For
1657 example, a match in host_lookup or the need for a host name in a host list.
1658 Now, if $sender_host_name is referenced and the host name has not yet been
1659 looked up, a lookup is performed. If the lookup fails, the variable remains
1660 empty, and $host_lookup_failed is set to "1".
1662 19. Added "eqi" as a case-independent comparison operator.
1664 20. The saslauthd authentication condition could segfault if neither service
1665 nor realm was specified.
1667 21. If an overflowing value such as "2048M" was set for message_size_limit, the
1668 error message that was logged was misleading, and incoming SMTP
1669 connections were dropped. The message is now more accurate, and temporary
1670 errors are given to SMTP connections.
1672 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
1673 (except RSET) with a 421 error, until QUIT is received. However, it was
1674 failing to send a response to QUIT.
1676 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
1677 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
1678 ACL is now run after the helo_try_verify_hosts code.
1680 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
1681 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
1682 case-independent, so other case variants are also recognized). Apparently
1683 some systems use these upper case variants.
1685 25. If more than two messages were waiting for the same host, and a transport
1686 filter was specified for the transport, Exim sent two messages over the
1687 same TCP/IP connection, and then failed with "socket operation on non-
1688 socket" when it tried to send the third.
1690 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
1692 27. The extern definition of crypt16() in expand.c was not being excluded when
1693 the OS had its own crypt16() function.
1695 28. Added bounce_return_body as a new option, and bounce_return_size_limit
1696 as a preferred synonym for return_size_limit, both as an option and as an
1699 29. Added LIBS=-liconv to OS/Makefile-OSF1.
1701 30. Changed the default configuration ACL to relax the local part checking rule
1702 for addresses that are not in any local domains. For these addresses,
1703 slashes and pipe symbols are allowed within local parts, but the sequence
1704 /../ is explicitly forbidden.
1706 31. SPA server authentication was not clearing the challenge buffer before
1709 32. log_message in a "warn" ACL statement was writing to the reject log as
1710 well as to the main log, which contradicts the documentation and doesn't
1711 seem right (because no rejection is happening). So I have stopped it.
1713 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
1714 However, I am unable to do any testing of this.
1716 34. Fixed an infelicity in the appendfile transport. When checking directories
1717 for a mailbox, to see if any needed to be created, it was accidentally
1718 using path names with one or more superfluous leading slashes; tracing
1719 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
1721 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
1722 contents are added to the log line that is written for every discarded
1723 recipient. (Previously a log_message setting was ignored.)
1725 36. The ${quote: operator now quotes the string if it is empty.
1727 37. The install script runs exim in order to find its version number. If for
1728 some reason other than non-existence or emptiness, which it checks, it
1729 could not run './exim', it was installing it with an empty version number,
1730 i.e. as "exim-". This error state is now caught, and the installation is
1733 38. An argument was missing from the function that creates an error message
1734 when Exim fails to connect to the socket for saslauthd authentication.
1735 This could cause Exim to crash, or give a corrupted message.
1737 39. Added isip, isip4, and isip6 to ${if conditions.
1739 40. The ACL variables $acl_xx are now saved with the message, and can be
1740 accessed later in routers, transports, and filters.
1742 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
1743 strings in the file are not string-expanded.
1745 42. If a MAIL command specified a SIZE value that was too large to fit into an
1746 int variable, the check against message_size_limit failed. Such values are
1747 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
1748 day this will have to be increased, but I don't think I want to be around
1749 when emails are that large.
1756 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
1757 iconv() is not standard in FreeBSD.
1759 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
1760 IPv6 enabled. The observed symptom was a segmentation fault on return from
1761 the function os_common_find_running_interfaces() in src/os.c.
1763 3. In the check_special_case() function in daemon.c I had used "errno" as an
1764 argument name, which causes warnings on some systems. This was basically a
1765 typo, since it was named "eno" in the comments!
1767 4. The code that waits for the clock to tick (at a resolution of some fraction
1768 of a second) so as to ensure message-id uniqueness was always waiting for
1769 at least one whole tick, when it could have waited for less. [This is
1770 almost certainly not relevant at current processor speeds, where it is
1771 unlikely to ever wait at all. But we try to future-proof.]
1773 5. The function that sleeps for a time interval that includes fractions of a
1774 second contained a race. It did not block SIGALRM between setting the
1775 timer, and suspending (a couple of lines later). If the interval was short
1776 and the sigsuspend() was delayed until after it had expired, the suspension
1777 never ended. On busy systems this could lead to processes getting stuck for
1780 6. Some uncommon configurations may cause a lookup to happen in a queue runner
1781 process, before it forks any delivery processes. The open lookup caching
1782 mechanism meant that the open file or database connection was passed into
1783 the delivery process. The problem was that delivery processes always tidy
1784 up cached lookup data. This could cause a problem for the next delivery
1785 process started by the queue runner, because the external queue runner
1786 process does not know about the closure. So the next delivery process
1787 still has data in the lookup cache. In the case of a file lookup, there was
1788 no problem because closing a file descriptor in a subprocess doesn't affect
1789 the parent. However, if the lookup was caching a connection to a database,
1790 the connection was closed, and the second delivery process was likely to
1791 see errors such as "PGSQL: query failed: server closed the connection
1792 unexpectedly". The problem has been fixed by closing all cached lookups
1793 in a queue runner before running a delivery process.
1795 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
1796 seem to have the "const" qualifier which it has on other OS. I've
1799 8. Change 4.21/2 was too strict. It is only if there are two authenticators
1800 *of the same type* (client or server) with the same public name that an
1801 error should be diagnosed.
1803 9. When Exim looked up a host name for an IP address, but failed to find the
1804 original IP address when looking up the host name (a safety check), it
1805 output the message "<ip address> does not match any IP for NULL", which was
1806 confusing, to say the least. The bug was that the host name should have
1807 appeared instead of "NULL".
1809 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
1810 user that is built into the binary, and the -C or -D options is used, root
1811 privilege is dropped before the configuration file is read. In addition,
1812 logging is switched to stderr instead of the normal log files. If the
1813 configuration then re-defines the Exim user, the unprivileged environment
1814 is probably not what is expected, so Exim logs a panic warning message (but
1817 However, if deliver_drop_privilege is set, the unprivileged state may well
1818 be exactly what is intended, so the warning has been cut out in that case,
1819 and Exim is allowed to try to write to its normal log files.
1825 1. smtp_return_error_details was not giving details for temporary sender
1826 or receiver verification errors.
1828 2. Diagnose a configuration error if two authenticators have the same public
1831 3. Exim used not to create the message log file for a message until the first
1832 delivery attempt. This could be confusing when incoming messages were held
1833 for policy or load reasons. The message log file is now created at the time
1834 the message is received, and an initial "Received" line is written to it.
1836 4. The automatically generated man page for command line options had a minor
1837 bug that caused no ill effects; however, a more serious problem was that
1838 the procedure for building the man page automatically didn't always
1839 operate. Consequently, release 4.20 contains an out-of-date version. This
1840 shouldn't happen again.
1842 5. When building Exim with embedded Perl support, the script that builds the
1843 Makefile was calling 'perl' to find its compile-time parameters, ignoring
1844 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
1846 6. The freeze_tell option was not being used for messages that were frozen on
1847 arrival, either by an ACL or by local_scan().
1849 7. Added the smtp_incomplete_transaction log selector.
1851 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
1852 was accepting AUTH without a new EHLO.
1854 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
1855 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
1858 10. Logging of TCP/IP connections (when configured) now happens in the main
1859 daemon process instead of the child process, so that the TCP/IP connection
1860 count is more accurate (but it can never be perfect).
1862 11. The use of "drop" in a nested ACL was not being handled correctly in the
1863 outer ACL. Now, if condition failure induced by the nested "drop" causes
1864 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
1865 or "require"), the connection is dropped.
1867 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
1868 that yield "discard" can now be used with an "accept" or a "discard" verb,
1869 but an error is generated for any others (because I can't see a useful way
1870 to define what should happen).
1872 13. When an ACL is read dynamically from a file (or anywhere else), the lines
1873 are now processed in the same way as lines in the Exim configuration file.
1874 In particular, continuation lines are supported.
1876 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
1878 15. Added -ti meaning -t -i.
1880 16. Check for letters, digits, hyphens, and dots in the names of dnslist
1881 domains, and warn by logging if others are found.
1883 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
1884 returned from GIFCONF when Exim is trying to find the list of interfaces on
1885 a host. The code in os.c has been modified to copy each ifreq to an aligned
1886 structure in all cases.
1888 Also, in some cases, the returned ifreq's were being copied to a 'struct
1889 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
1890 means the last couple of bytes of an IPv6 address could be chopped if the
1891 ifreq contained only a normal sockaddr (14 bytes storage).
1893 18. Named domain lists were not supported in the hosts_treat_as_local option.
1894 An entry such as +xxxx was not recognized, and was treated as a literal
1897 19. Ensure that header lines added by a DATA ACL are included in the reject log
1898 if the ACL subsequently rejects the message.
1900 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
1901 MD5 (which is deprecated).
1903 21. When testing a filter file using -bf, Exim was writing a message when it
1904 took the sender from a "From " line in the message, but it was not doing so
1905 when it took $return_path from a Return-Path: header line. It now does.
1907 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
1908 with a valid header line field name (a series of printing characters
1909 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
1911 23. Changed "disc" in the source to "disk" to conform to the documentation and
1912 the book and for uniformity.
1914 24. Ignore Sendmail's -Ooption=value command line item.
1916 25. When execve() failed while trying to run a command in a pipe transport,
1917 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
1918 could be confused with a return value of 69 from the command itself. This
1919 has been changed to 127, the value the shell returns if it is asked to run
1920 a non-existent command. The wording for the related log line suggests a
1921 non-existent command as the problem.
1923 26. If received_header_text expands to an empty string, do not add a Received:
1924 header line to the message. (Well, it adds a token one on the spool, but
1925 marks it "old" so that it doesn't get used or transmitted.)
1927 27. Installed eximstats 1.28 (addition of -nt option).
1929 28. There was no check for failure on the call to getsockname() in the daemon
1930 code. This can fail if there is a shortage of resources on the system, with
1931 ENOMEM, for example. A temporary error is now given on failure.
1933 29. Contrary to the C standard, it seems that in some environments, the
1934 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
1935 program. Exim now does this explicitly; it affects the formatting of
1936 timestamps using strftime().
1938 30. If exiqsumm was given junk data, it threw up some uninitialized variable
1939 complaints. I've now initialized all the variables, to avoid this.
1941 32. Header lines added by a system filter were not being "seen" during
1942 transport-time rewrites.
1944 33. The info_callback() function passed to OpenSSL is set up with type void
1945 (*)(SSL *, int, int), as described somewhere. However, when calling the
1946 function (actually a macro) that sets it up, the type void(*)() is
1947 expected. I've put in a cast to prevent warnings from picky compilers.
1949 34. If a DNS black list lookup found a CNAME record, but there were no A
1950 records associated with the domain it pointed at, Exim crashed.
1952 35. If a DNS black list lookup returned more than one A record, Exim ignored
1953 all but the first. It now scans all returned addresses if a particular IP
1954 value is being sought. In this situation, the contents of the
1955 $dnslist_value variable are a list of all the addresses, separated by a
1958 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
1959 to accept a host name and all its aliases if the forward lookup for any of
1960 them yielded the IP address of the incoming connection. Now it accepts only
1961 those names whose forward lookup yields the correct IP address. Any other
1962 names are discarded. This closes a loophole whereby a rogue DNS
1963 administrator could create reverse DNS records to break through a
1964 wildcarded host restriction in an ACL.
1966 37. If a user filter or a system filter that ran in a subprocess used any of
1967 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
1968 the wrong values were passed to the pipe command ($thisaddress had the
1969 value of $0, $0 had the value of $1, etc). This bug was introduced by
1970 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
1972 38. Improved the line breaking for long SMTP error messages from ACLs.
1973 Previously, if there was no break point between 40 and 75 characters, Exim
1974 left the rest of the message alone. Two changes have been made: (a) I've
1975 reduced the minimum length to 35 characters; (b) if it can't find a break
1976 point between 35 and 75 characters, it looks ahead and uses the first one
1977 that it finds. This may give the occasional overlong line, but at least the
1978 remaining text gets split now.
1980 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
1981 file descriptors might be low, and that setting 1000 would always raise it.
1982 It turns out that in some environments, the limit is already over 1000 and
1983 that lowering it causes trouble. So now Exim takes care not to decrease it.
1985 40. When delivering a message, the value of $return_path is set to $sender_
1986 address at the start of routing (routers may change the value). By an
1987 oversight, this default was not being set up when an address was tested by
1988 -bt or -bv, which affected the outcome if any router or filter referred to
1991 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
1992 log only when "message" or "log_message" are set. However, if one of the
1993 conditions was an address verification, or a call to a nested ACL, the
1994 messages generated by the underlying test were being passed through. This
1995 no longer happens. The underlying message is available in $acl_verify_
1996 message for both "message" and "log_message" expansions, so it can be
1997 passed through if needed.
1999 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2000 new expansion $bh_ to give the encoded byte string without charset
2001 translation. Translation happens only if iconv() is available; HAVE_ICONV
2002 indicates this at build time. HEADERS_CHARSET gives the charset to
2003 translate to; headers_charset can change it in the configuration, and
2004 "headers charset" can change it in an individual filter file.
2006 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
2007 that creates RFC 2047 encoded "words" labels them as that charset instead
2008 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2009 expansion operator; (ii) when Exim creates a From: line for a local
2010 message; (iii) when a header line is rewritten to include a "phrase" part.
2012 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2013 buggy, causing it to skip the first lines of messages whose message ID
2014 ended in 'D'. This would not have bitten before Exim release 4.14, because
2015 message IDs were unlikely to end in 'D' before then. The effect was to have
2016 incorrect size information for certain domains.
2018 45. #include "config.h" was missing at the start of the crypt16.c module. This
2019 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2022 46. If there was a timeout during a "random" callout check, Exim treated it as
2023 a failure of the random address, and carried on sending RSET and the real
2024 address. If the delay was just some slowness somewhere, the response to the
2025 original RCPT would be taken as a response to RSET and so on, causing
2026 mayhem of various kinds.
2028 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2029 when I implemented it. It didn't allow for the fact that some option values
2030 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2031 the right test for positive values.
2033 48. Domain names in DNS records are case-independent. Exim always looks them up
2034 in lower case. Some resolvers return domain names in exactly the case they
2035 appear in the zone file, that is, they may contain uppercase letters. Not
2036 all resolvers do this - some return always lower case. Exim was treating a
2037 change of case by a resolver as a change of domain, similar to a widening
2038 of a domain abbreviation. This triggered its re-routing code and so it was
2039 trying to route what was effectively the same domain again. This normally
2040 caused routing to fail (because the router wouldn't handle the domain
2041 twice). Now Exim checks for this case specially, and just changes the
2042 casing of the domain that it ultimately uses when it transmits the message
2045 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
2048 50. If a filter generated a file delivery with a non-absolute name (possible if
2049 no home directory exists for the router), the forbid_file option was not
2052 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
2053 the existing equality matching.
2055 52. Exim was using ints instead of ino_t variables in some places where it was
2056 dealing with inode numbers.
2058 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
2059 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
2060 called TMPDIR, and if it finds it is different, it changes its value.
2062 54. The smtp_printf() function is now made available to local_scan() so
2063 additional output lines can be written before returning. There is also an
2064 smtp_fflush() function to enable the detection of a dropped connection.
2065 The variables smtp_input and smtp_batched_input are exported to
2068 55. Changed the default runtime configuration: the message "Unknown user"
2069 has been removed from the ACL, and instead placed on the localuser router,
2070 using the cannot_route_message feature. This means that any verification
2071 failures that generate their own messages won't get overridden. Similarly,
2072 the "Unrouteable address" message that was in the ACL for unverifiable
2073 relay addresses has also been removed.
2075 56. Added hosts_avoid_esmtp to the smtp transport.
2077 57. The exicyclog script was not checking for the esoteric option
2078 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
2079 will work only if exicyclog is run under the appropriate euid.
2081 58. Following a discussion on the list, the rules by which Exim recognises line
2082 endings on incoming messages have been changed. The -dropcr and drop_cr
2083 options are now no-ops, retained only for backwards compatibility. The
2084 following line terminators are recognized: LF CRLF CR. However, special
2085 processing applies to CR:
2087 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
2088 nor a local message in the state where . is a terminator.
2090 (ii) If a bare CR is encountered in a header line, an extra space is added
2091 after the line terminator so as not to end the header. The reasoning
2092 behind this is that bare CRs in header lines are most likely either
2093 to be mistakes, or people trying to play silly games.
2095 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
2096 was being incorrectly given as 18 bytes larger than it should have been.
2097 This is a VOB (very old bug).
2099 60. This may never have affected anything current, but just in case it has:
2100 When the local host is found other than at the start of a list of hosts,
2101 the local host, those with the same MX, and any that follow, are discarded.
2102 When the list in question was part of a longer list of hosts, the following
2103 hosts (not currently being processed) were also being discarded. This no
2104 longer happens. I'm not sure if this situation could ever has previously
2107 61. Added the "/MX" feature to lists of hosts in the manualroute and query
2110 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
2111 header. This is something that is recommended in a new Internet Draft, and
2112 is something that is documented as being done by Sendmail. There are two
2113 possible values. For messages generated by the autoreply transport, Exim
2116 Auto-Submitted: auto-replied
2118 whereas for all other generated messages (e.g. bounces) it adds
2120 Auto-Submitted: auto-generated
2122 63. The "personal" condition in filters now includes a test for the
2123 Auto-Submitted: header. If it contains the string "auto-" the message it
2124 not considered personal.
2126 64. Added rcpt_include_affixes as a generic transport option.
2128 65. Added queue_only_override (default true).
2130 66. Added the syslog_duplication option.
2132 67. If what should have been the first header line of a message consisted of
2133 a space followed by a colon, Exim was mis-interpreting it as a header line.
2134 It isn't of course - it is syntactically invalid and should therefore be
2135 treated as the start of the message body. The misbehaviour could have
2136 caused a number of strange effects, including loss of data in subsequent
2137 header lines, and spool format errors.
2139 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
2140 client host had authenticated. This control can now be exercised by an ACL
2141 for more flexibility.
2143 69. By default, callouts do not happen when testing with -bh. There is now a
2144 variant, -bhc, which does actually run the callout code, including
2145 consulting and updating the callout cache.
2147 70. Added support for saslauthd authentication, courtesy of Alexander
2150 71. If statvfs() failed on the spool or log directories while checking their
2151 size for availability, Exim confusingly gave the error "space shortage".
2152 Furthermore, in debugging mode it crashed with a floating point exception.
2153 These checks are done if check_{spool,log}_{space,inodes} are set, and when
2154 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
2155 serious problem, Exim now writes to the main and panic logs when this
2156 happens, with details of the failure. It then refuses to accept the
2157 incoming message, giving the message "spool directory problem" or "log
2158 directory problem" with a 421 code for SMTP messages.
2160 72. When Exim is about to re-exec itself, it ensures that the file descriptors
2161 0, 1, and 2 exist, because some OS complain for execs without them (see
2162 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
2163 descriptors. However, the code omitted to check that the open succeeded,
2164 causing mysterious errors if for some reason the permissions on /dev/null
2165 got screwed. Now Exim writes a message to the main and panic logs, and
2166 bombs out if it can't open /dev/null.
2168 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
2169 interact so that it is all more flexible. It is supposed to remain
2170 backwards compatible. Also added extra_local_interfaces.
2172 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
2173 to bomb out with an assertion failure - to the client this appears as a
2174 connection drop. This problem occurs in the part of the code that was taken
2175 from the Samba project. Fortunately, the assertion is in a very simple
2176 function, so I have fixed this by reproducing the function inline in the
2177 one place where it is called, and arranging for authentication to fail
2178 instead of killing the process with assert().
2180 75. The SPA client code was not working when the server requested OEM rather
2181 than Unicode encoding.
2183 76. Added code to make require_files with a specific uid setting more usable in
2184 the case where statting the file as root fails - usually a non-root-mounted
2185 NFS file system. When this happens and the failure is EACCES, Exim now
2186 forks a subprocess and does the per-uid checking as the relevant uid.
2188 77. Added process_log_path.
2190 78. If log_file_path was not explicitly set, a setting of check_log_space or
2191 check_log_inodes was ignored.
2193 79. If a space check for the spool or log partitions fails, the incident is now
2194 logged. Of course, in the latter case the data may get lost...
2196 80. Added the %p formatting code to string_format() so that it can be used to
2197 print addresses in debug_print(). Adjusted all the address printing in the
2198 debugging in store.c to use %p rather than %d.
2200 81. There was a concern that a line of code in smtp_in.c could overflow a
2201 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
2202 initially expressed, the concern was not well-founded, because trailing
2203 spaces are removed early. However, if the trailing spaces were followed by
2204 a NULL, they did not get removed, so the overflow was possible. Two fixes
2207 (a) I re-wrote the offending code in a cleaner fashion.
2208 (b) If an incoming SMTP command contains a NULL character, it is rejected
2211 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
2212 runs initgroups(), so that if the Exim user is in any additional groups,
2213 they will be used during message reception.
2219 The change log for 4.20 and earlier releases has been archived.