1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.59 2004/12/29 10:55:58 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 1. Minor wording change to the doc/README.SIEVE file.
11 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
12 computation of the current number of files was incorrect.
14 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
15 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
18 4. Give more explanation in the error message when the command for a transport
19 filter fails to execute.
21 5. There are several places where Exim runs a non-Exim command in a
22 subprocess. The SIGUSR1 signal should be disabled for these processes. This
23 was being done only for the command run by the queryprogram router. It is
24 now done for all such subprocesses. The other cases are: ${run, transport
25 filters, and the commands run by the lmtp and pipe transports.
27 6. Added CONFIGURE_GROUP build-time option.
29 7. Some older OS have a limit of 256 on the maximum number of file
30 descriptors. Exim was using setrlimit() to set 1000 as a large value
31 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
32 systems. I've change it so that if it can't get 1000, it tries for 256.
34 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
35 was an oversight, and furthermore, ever since the addition of extra
36 controls (e.g. 4.43/32), the checks on when to allow different forms of
37 "control" were broken. There should now be diagnostics for all cases when a
38 control that does not make sense is encountered.
40 9. Added the /retain_sender option to "control=submission".
42 10. $recipients is now available in the predata ACL (oversight).
44 11. Tidy the search cache before the fork to do a delivery from a message
45 received from the command line. Otherwise the child will trigger a lookup
46 failure and thereby defer the delivery if it tries to use (for example) a
47 cached ldap connection that the parent has called unbind on.
49 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
50 of $address_data from the recipient verification was clobbered by the
53 13. The value of address_data from a sender verification is now available in
54 $sender_address_data in subsequent conditions in the ACL statement.
56 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
58 15. Added a new option "connect=<time>" to callout options, to set a different
61 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
62 was its contents. (It was OK if the option was not defined at all.)
64 17. A "Completed" log line is now written for messages that are removed from
65 the spool by the -Mrm option.
67 18. New variables $sender_verify_failure and $recipient_verify_failure contain
68 information about exactly what failed.
70 19. Added -dd to debug only the daemon process.
72 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
73 handles timeouts, both on the server side and network timeouts. Renamed the
74 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
76 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
77 It is now set to "smtps".
79 22. $host_address is now set to the target address during the checking of
82 23. When checking ignore_target_hosts for an ipliteral router, no host name was
83 being passed; this would have caused $sender_host_name to have been used if
84 matching the list had actually called for a host name (not very likely,
85 since this list is usually IP addresses). A host name is now passed as
88 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
89 code that specifies a non-restarting handler (typically sigaction() in
90 modern systems) in an attempt to fix a rare and obscure crash bug.
92 25. Narrowed the window for a race in the daemon that could cause it to ignore
93 SIGCHLD signals. This is not a major problem, because they are used only to
94 wake it up if nothing else does.
96 26. A malformed maildirsize file could cause Exim to calculate negative values
97 for the mailbox size or file count. Odd effects could occur as a result.
98 The maildirsize information is now recalculated if the size or filecount
101 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
102 support for a long time. Removed HAVE_SYS_VFS_H.
104 28. Installed the latest version of exipick from John Jetmore.
106 29. In an address list, if the pattern was not a regular expression, an empty
107 subject address (from a bounce message) matched only if the pattern was an
108 empty string. Non-empty patterns were not even tested. This was the wrong
109 because it is perfectly reasonable to use an empty address as part of a
110 database query. An empty address is now tested by patterns that are
111 lookups. However, all the other forms of pattern expect the subject to
112 contain a local part and a domain, and therefore, for them, an empty
113 address still always fails if the pattern is not itself empty.
115 30. Exim went into a mad DNS loop when attempting to do a callout where the
116 host was specified on an smtp transport, and looking it up yielded more
119 31. Re-factored the code for checking spool and log partition space into a
120 function that finds that data and another that does the check. The former
121 is then used to implement four new variables: $spool_space, $log_space,
122 $spool_inodes, and $log_inodes.
124 32. The RFC2047 encoding function was originally intended for short strings
125 such as real names; it was not keeping to the 75-character limit for
126 encoded words that the RFC imposes. It now respects the limit, and
127 generates multiple encoded words if necessary. To be on the safe side, I
128 have increased the buffer size for the ${rfc2047: expansion operator from
131 33. It is now permitted to omit both strings after an "if" condition; if the
132 condition is true, the result is "true". As before, when the second string
133 is omitted, a false condition yields an empty string. This makes it less
134 cumbersome to write custom ACL and router conditions.
136 34. Failure to deliver a bounce message always caused it to be frozen, even if
137 there was an errors_to setting on the router. The errors_to setting is now
140 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
141 canonical form (fully expanded) before being placed in
142 $sender_host_address.
144 36. The table in the code that translates DNS record types into text (T_A to
145 "A" for instance) was missing entries for NS and CNAME. It is just possible
146 that this could have caused confusion if both these types were looked up
147 for the same domain, because the text type is used as part of Exim's
148 per-process caching. But the chance of anyone hitting this buglet seems
151 37. The dnsdb lookup has been extended in a number of ways.
153 (1) There is a new type, "zns", which walks up the domain tree until it
154 finds some nameserver records. It should be used with care.
156 (2) There is a new type, "mxh", which is like "mx" except that it returns
157 just the host names, not the priorities.
159 (3) It is now possible to give a list of domains (or IP addresses) to be
160 looked up. The behaviour when one of the lookups defers can be
161 controlled by a keyword.
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.
175 41. Include certificate and key file names in error message when GnuTLS fails
176 to set them up, because the GnuTLS error message doesn't include the name
177 of the failing file when there is a problem reading it.
179 42. Allow both -bf and -bF in the same test run.
181 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
183 44. The "Exiscan patch" is now merged into the mainline Exim source.
185 45. Sometimes the final signoff response after QUIT could fail to get
186 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
187 < 0 before doing a fflush(). This bug looks as though it goes back to the
188 introduction of TLS in release 3.20, but "sometimes" must have been rare
189 because the tests only now provoked it.
191 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
192 (this can affect the format of dates).
194 47. exim_tidydb, when checking for the continued existence of a message for
195 which it has found a message-specific retry record, was not finding
196 messages that were in split spool directories. Consequently, it was
197 deleting retry records that should have stayed in existence.
199 48. Steve fixed some bugs in eximstats.
201 49. The SPA authentication driver was not abandoning authentication and moving
202 on to the next authenticator when an expansion was forced to fail,
203 contradicting the general specification for all authenticators. Instead it
204 was generating a temporary error. It now behaves as specified.
206 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
207 (the order specifies the preference for clients). The order is now AES256,
208 AES128, 3DES, ARCFOUR128.
210 51. Small patch to Sieve code - explicitly set From: when generating an
213 52. Exim crashed if a remote delivery caused a very long error message to be
214 recorded - for instance if somebody sent an entire SpamAssassin report back
215 as a large number of 550 error lines. This bug was coincidentally fixed by
216 increasing the size of one of Exim's internal buffers (big_buffer) that
217 happened as part of the Exiscan merge. However, to be on the safe side, I
218 have made the code more robust (and fixed the comments that describe what
221 53. Now that there can be additional text after "Completed" in log lines (if
222 the queue_time_overall log selector is set), a one-byte patch to exigrep
223 was needed to allow it to recognize "Completed" as not the last thing in
226 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
227 patch that reportedly fixes this has been added. I am not expert enough to
228 create a test for it. This is what the patch creator wrote:
230 "I found a little strange behaviour of ldap code when working with
231 Windows 2003 AD Domain, where users was placed in more than one
232 Organization Units. When I tried to give exim partial DN, the exit code
233 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
234 But simultaneously result of request was absolutely normal ldap result,
235 so I produce this patch..."
237 55. Some experimental protocols are using DNS PTR records for new purposes. The
238 keys for these records are domain names, not reversed IP addresses. The
239 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
240 leaves it alone. Component reversal etc. now happens only for IP addresses.
242 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
244 57. Double the size of the debug message buffer (to 2048) so that more of very
245 long debug lines gets shown.
247 58. The exicyclog utility now does better if the number of log files to keep
248 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
250 59. Two changes related to the smtp_active_hostname option:
252 (1) $smtp_active_hostname is now available as a variable.
253 (2) The default for smtp_banner uses $smtp_active_hostname instead
254 of $primary_hostname.
256 60. The host_aton() function is supposed to be passed a string that is known
257 to be a valid IP address. However, in the case of IPv6 addresses, it was
258 not checking this. This is a hostage to fortune. Exim now panics and dies
259 if the condition is not met. A case was found where this could be provoked
260 from a dnsdb PTR lookup with an IPv6 address that had more than 8
261 components; fortuitously, this particular loophole had already been fixed
262 by change 4.50/55 above.
264 If there are any other similar loopholes, the new check in host_aton()
265 itself should stop them being exploited. The report I received stated that
266 data on the command line could provoke the exploit when Exim was running as
267 exim, but did not say which command line option was involved. All I could
268 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
271 61. There was a buffer overflow vulnerability in the SPA authentication code
272 (which came originally from the Samba project). I have added a test to the
273 spa_base64_to_bits() function which I hope fixes it.
279 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
280 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
281 Now it returns an integer. A number of calls were still expecting a T/F
282 return. Fortuitously, in all cases, the tests worked in OK situations,
283 which is the norm. However, things would have gone wrong on any write
284 failures on the smtp file descriptor. This function is used when sending
285 messages over SMTP and also when doing verify callouts.
287 2. When Exim is called to do synchronous delivery of a locally submitted
288 message (the -odf or -odi options), it no longer closes stderr before doing
291 3. Implemented the mua_wrapper option.
293 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
295 5. Implemented the functions header_remove(), header_testname(),
296 header_add_at_position(), and receive_remove_recipient(), and exported them
299 6. If an ACL "warn" statement specified the addition of headers, Exim already
300 inserted X-ACL-Warn: at the start if there was no header name. However, it
301 was not making this test for the second and subsequent header lines if
302 there were newlines in the string. This meant that an invalid header could
303 be inserted if Exim was badly configured.
305 7. Allow an ACL "warn" statement to add header lines at the start or after all
306 the Received: headers, as well as at the end.
308 8. Added the rcpt_4xx retry error code.
310 9. Added postmaster_mailfrom=xxx to callout verification option.
312 10. Added mailfrom=xxxx to the callout verification option, for verify=
315 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
316 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
318 12. Inserted some casts to stop certain compilers warning when using pointer
319 differences as field lengths or precisions in printf-type calls (mostly
320 affecting debugging statements).
322 13. Added optional readline() support for -be (dynamically loaded).
324 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
325 same clock tick as a message's arrival, so that its received time was the
326 same as the "first fail" time on the retry record, and that message
327 remained on the queue past the ultimate address timeout, every queue runner
328 would try a delivery (because it was past the ultimate address timeout) but
329 after another failure, the ultimate address timeout, which should have then
330 bounced the address, did not kick in. This was a "< instead of <=" error;
331 in most cases the first failure would have been in the next clock tick
332 after the received time, and all would be well.
334 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
335 being recognized when the domain list was tested by the match_domain
336 condition in an expansion string.
338 16. Added the ${str2b64: operator.
340 17. Exim was always calling setrlimit() to set a large limit for the number of
341 processes, without checking whether the existing limit was already
342 adequate. (It did check for the limit on file descriptors.) Furthermore,
343 errors from getrlimit() and setrlimit() were being ignored. Now they are
344 logged to the main and panic logs, but Exim does carry on, to try to do its
345 job under whatever limits there are.
347 18. Imported PCRE 5.0.
349 19. Trivial typo in log message " temporarily refused connection" (the leading
352 20. If the log selector return_path_on_delivery was set and an address was
353 redirected to /dev/null, the delivery process crashed because it assumed
354 that a return path would always be set for a "successful" delivery. In this
355 case, the whole delivery is bypassed as an optimization, and therefore no
358 21. Internal re-arrangement: the function for sending a challenge and reading
359 a response while authentication was assuming a zero-terminated challenge
360 string. It's now changed to take a pointer and a length, to allow for
361 binary data in such strings.
363 22. Added the cyrus_sasl authenticator (code supplied by MBM).
365 23. Exim was not respecting finduser_retries when seeking the login of the
366 uid under which it was called; it was always trying 10 times. (The default
367 setting of finduser_retries is zero.) Also, it was sleeping after the final
368 failure, which is pointless.
370 24. Implemented tls_on_connect_ports.
372 25. Implemented acl_smtp_predata.
374 26. If the domain in control=submission is set empty, Exim assumes that the
375 authenticated id is a complete email address when it generates From: or
376 Sender: header lines.
378 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
379 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
380 chown and chgrp in /bin and hostname in /usr/bin.
382 28. Exim was keeping the "process log" file open after each use, just as it
383 does for the main log. This opens the possibility of it remaining open for
384 long periods when the USR1 signal hits a daemon. Occasional processlog
385 errors were reported, that could have been caused by this. Anyway, it seems
386 much more sensible not to leave this file open at all, so that is what now
389 29. The long-running daemon process does not normally write to the log once it
390 has entered its main loop, and it closes the log before doing so. This is
391 so that log files can straightforwardly be renamed and moved. However,
392 there are a couple of unusual error situations where the daemon does write
393 log entries, and I had neglected to close the log afterwards.
395 30. The text of an SMTP error response that was received during a remote
396 delivery was being truncated at 512 bytes. This is too short for some of
397 the long messages that one sometimes sees. I've increased the limit to
400 31. It is now possible to make retry rules that apply only when a message has a
401 specific sender, in particular, an empty sender.
403 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
404 it possible to be selective about when SMTP synchronization is enforced.
406 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
408 32. Implemented hosts_connection_nolog.
410 33. Added an ACL for QUIT.
412 34. Setting "delay_warning=" to disable warnings was not working; it gave a
415 35. Added mailbox_size and mailbox_filecount to appendfile.
417 36. Added control = no_multiline_responses to ACLs.
419 37. There was a bug in the logic of the code that waits for the clock to tick
420 in the case where the clock went backwards by a substantial amount such
421 that the microsecond fraction of "now" was more than the microsecond
422 fraction of "then" (but the whole seconds number was less).
424 38. Added support for the libradius Radius client library this is found on
425 FreeBSD (previously only the radiusclient library was supported).
431 1. When certain lookups returned multiple values in the form name=value, the
432 quoting of the values was not always being done properly. Specifically:
433 (a) If the value started with a double quote, but contained no whitespace,
435 (b) If the value contained whitespace other than a space character (i.e.
436 tabs or newlines or carriage returns) it was not quoted.
437 This fix has been applied to the mysql and pgsql lookups by writing a
438 separate quoting function and calling it from the lookup code. The fix
439 should probably also be applied to nisplus, ibase and oracle lookups, but
440 since I cannot test any of those, I have not disturbed their existing code.
442 2. A hit in the callout cache for a specific address caused a log line with no
443 reason for rejecting RCPT. Now it says "Previous (cached) callout
444 verification failure".
446 3. There was an off-by-one bug in the queryprogram router. An over-long
447 return line was truncated at 256 instead of 255 characters, thereby
448 overflowing its buffer with the terminating zero. As well as fixing this, I
449 have increased the buffer size to 1024 (and made a note to document this).
451 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
452 when Exim is waiting for an SMTP response from a remote server, Exim
453 restarts its select() call on the socket, thereby resetting its timeout.
454 This is not a problem when such interrupts are rare. Somebody set up a cron
455 job to run exiwhat every 2 minutes, which is less than the normal select()
456 timeout (5 or 10 minutes). This meant that the select() timeout never
457 kicked in because it was always reset. I have fixed this by comparing the
458 time when an interrupt arrives with the time at the start of the first call
459 to select(). If more time than the timeout has elapsed, the interrupt is
460 treated as a timeout.
462 5. Some internal re-factoring in preparation for the addition of Sieve
463 extensions (by MH). In particular, the "personal" test is moved to a
464 separate function, and given an option for scanning Cc: and Bcc: (which is
465 not set for Exim filters).
467 6. When Exim created an email address using the login of the caller as the
468 local part (e.g. when creating a From: or Sender: header line), it was not
469 quoting the local part when it contained special characters such as @.
471 7. Installed new OpenBSD configuration files.
473 8. Reworded some messages for syntax errors in "and" and "or" conditions to
474 try to make them clearer.
476 9. Callout options, other than the timeout value, were being ignored when
477 verifying sender addresses in header lines. For example, when using
479 verify = header_sender/callout=no_cache
481 the cache was (incorrectly) being used.
483 10. Added a missing instance of ${EXE} to the exim_install script; this affects
484 only the Cygwin environment.
486 11. When return_path_on_delivery was set as a log selector, if different remote
487 addresses in the same message used different return paths and parallel
488 remote delivery occurred, the wrong values would sometimes be logged.
489 (Whenever a remote delivery process finished, the return path value from
490 the most recently started remote delivery process was logged.)
492 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
493 lines when AUTH and/or TLS are in use. This is the "received protocol"
494 field. Exim used to use "asmtp" for authenticated SMTP, without any
495 indication (in the protocol name) for TLS use. Now it follows the RFC and
496 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
497 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
498 names appear in log lines as well as in Received: header lines.
500 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
501 extensions, and comparison tests, and to fix some bugs.
503 14. Changes to the "personal" filter test:
505 (1) The test was buggy in that it was just doing the equivalent of
506 "contains" tests on header lines. For example, if a user's address was
507 anne@some.where, the "personal" test would incorrectly be true for
509 To: susanne@some.where
511 This test is now done by extracting each address from the header in turn,
512 and checking the entire address. Other tests that are part of "personal"
513 are now done using regular expressions (for example, to check local parts
514 of addresses in From: header lines).
516 (2) The list of non-personal local parts in From: addresses has been
517 extended to include "listserv", "majordomo", "*-request", and "owner-*",
518 taken from the Sieve specification recommendations.
520 (3) If the message contains any header line starting with "List-" it is
521 treated as non-personal.
523 (4) The test for "circular" in the Subject: header line has been removed
524 because it now seems ill-conceived.
526 15. Minor typos in src/EDITME comments corrected.
528 16. Installed latest exipick from John Jetmore.
530 17. If headers_add on a router specified a text string that was too long for
531 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
532 of string_sprintf() is now avoided.
534 18. $message_body_size was not set (it was always zero) when running the DATA
535 ACL and the local_scan() function.
537 19. For the "mail" command in an Exim filter, no default was being set for
538 the once_repeat time, causing a random time value to be used if "once" was
539 specified. (If the value happened to be <= 0, no repeat happened.) The
540 default is now 0s, meaning "never repeat". The "vacation" command was OK
541 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
542 (I found it when inspecting the code).
544 20. There is now an overall timeout for performing a callout verification. It
545 defaults to 4 times the callout timeout, which applies to individual SMTP
546 commands during the callout. The overall timeout applies when there is more
547 than one host that can be tried. The timeout is checked before trying the
548 next host. This prevents very long delays if there are a large number of
549 hosts and all are timing out (e.g. when the network connections are timing
550 out). The value of the overall timeout can be changed by specifying an
551 additional sub-option for "callout", called "maxwait". For example:
553 verify = sender/callout=5s,maxwait=20s
555 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
556 to the end before writing, but this should make it even safer).
558 22. Exim was forgetting that it had advertised PIPELINING for the second and
559 subsequent messages on an SMTP connection. It was also not resetting its
560 memory on STARTTLS and an internal HELO.
562 23. When Exim logs an SMTP synchronization error within a session, it now
563 records whether PIPELINING has been advertised or not.
565 24. Added 3 instances of "(long int)" casts to time_t variables that were being
566 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
567 rather than long int.
569 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
571 26. Added the never_mail option to autoreply.
577 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
578 crash if the getsockname() call failed; this can happen if a connection is
579 closed very soon after it is established. The problem was simply in the
580 order in which certain operations were done, causing Exim to try to write
581 to the SMTP stream before it had set up the file descriptor. The bug has
582 been fixed by making things happen in the correct order.
588 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
589 before the connection was closed, thus losing the rejection response.
591 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
592 some early Solaris releases, but causes trouble in current releases where
593 socklen_t is defined.
595 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
598 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
601 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
602 that releasing the top of it at the end releases what was used for sub-
603 expansions (unless the block got too big). However, discard this block if
604 the first thing is a variable or header, so that we can use its block when
605 it is dynamic (useful for very large $message_headers, for example).
607 6. Lookups now cache *every* query, not just the most recent. A new, separate
608 store pool is used for this. It can be recovered when all lookup caches are
609 flushed. Lookups now release memory at the end of their result strings.
610 This has involved some general refactoring of the lookup sources.
612 7. Some code has been added to the store_xxx() functions to reduce the amount
613 of flapping under certain conditions.
615 8. log_incoming_interface used to affect only the <= reception log lines. Now
616 it causes the local interface and port to be added to several more SMTP log
617 lines, for example "SMTP connection from", and rejection lines.
619 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
621 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
623 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
624 could be overwritten at the end of the current message (or the start of a
625 new message if it was set in a HELO ACL). The value is now preserved for
626 the duration of the SMTP connection.
628 12. If a transport had a headers_rewrite setting, and a matching header line
629 contained an unqualified address, that address was qualified, even if it
630 did not match any rewriting rules. The underlying bug was that the values
631 of the flags that permit the existence of unqualified sender and recipient
632 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
633 non-local messages, and by -bnq for local messages) were not being
634 preserved with the message after it was received.
636 13. When Exim was logging an SMTP synchronization error, it could sometimes log
637 "next input=" as part of the text comprising the host identity instead of
638 the correct text. The code was using the same buffer for two different
639 strings. However, depending on which order the printing function evaluated
640 its arguments, the bug did not always show up. Under Linux, for example, my
641 test suite worked just fine.
643 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
644 doesn't work with some older versions of Perl. It has been changed to "my",
645 which in any case is probably the better facility to use.
647 15. A really picky compiler found some instances of statements for creating
648 error messages that either had too many or two few arguments for the format
651 16. The size of the buffer for calls to the DNS resolver has been increased
652 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
653 for addresses that have a lot of PTR records. This alleviates a problem; it
654 does not fully solve it.
656 17. A dnsdb lookup for PTR records that receives more data than will fit in the
657 buffer now truncates the list and logs the incident, which is the same
658 action as happens when Exim is looking up a host name and its aliases.
659 Previously in this situation something unpredictable would happen;
660 sometimes it was "internal error: store_reset failed".
662 18. If a server dropped the connection unexpectedly when an Exim client was
663 using GnuTLS and trying to read a response, the client delivery process
664 crashed while trying to generate an error log message.
666 19. If a "warn" verb in an ACL added multiple headers to a message in a single
669 warn message = H1: something\nH2: something
671 the text was added as a single header line from Exim's point of view
672 though it ended up OK in the delivered message. However, searching for the
673 second and subsequent header lines using $h_h2: did not work. This has been
674 fixed. Similarly, if a system filter added multiple headers in this way,
675 the routers could not see them.
677 20. Expanded the error message when iplsearch is called with an invalid key to
678 suggest using net-iplsearch in a host list.
680 21. When running tests using -bh, any delays imposed by "delay" modifiers in
681 ACLs are no longer actually imposed (and a message to that effect is
684 22. If a "gecos" field in a passwd entry contained escaped characters, in
685 particular, if it contained a \" sequence, Exim got it wrong when building
686 a From: or a Sender: header from that name. A second bug also caused
687 incorrect handling when an unquoted " was present following a character
690 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
691 was not being matched caselessly.
693 24. Arranged for all hyphens in the exim.8 source to be escaped with
696 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
697 properly. Recipient callout cache records were still being keyed to include
698 the sender, even when use_sender was set false. This led to far more
699 callouts that were necessary. The sender is no longer included in the key
700 when use_sender is false.
702 26. Added "control = submission" modifier to ACLs.
704 27. Added the ${base62d: operator to decode base 62 numbers.
706 28. dnsdb lookups can now access SRV records.
708 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
709 the configuration file.
711 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
712 (-v) mode. This makes the output for a verbose queue run more intelligible.
714 31. Added a use_postmaster feature to recipient callouts.
716 32. Added the $body_zerocount variable, containing the number of binary zero
717 bytes in the message body.
719 33. The time of last modification of the "new" subdirectory is now used as the
720 "mailbox time last read" when there is a quota error for a maildir
723 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
725 35. Added +ignore_unknown as a special item in host lists.
727 36. Code for decoding IPv6 addresses in host lists is now included, even if
728 IPv6 support is not being compiled. This fixes a bug in which an IPv6
729 address was recognized as an IP address, but was then not correctly decoded
730 into binary, causing unexpected and incorrect effects when compared with
737 1. Very minor rewording of debugging text in manualroute to say "list of
738 hosts" instead of "hostlist".
740 2. If verify=header_syntax was set, and a header line with an unqualified
741 address (no domain) and a large number of spaces between the end of the
742 name and the colon was received, the reception process suffered a buffer
743 overflow, and (when I tested it) crashed. This was caused by some obsolete
744 code that should have been removed. The fix is to remove it!
746 3. When running in the test harness, delay a bit after writing a bounce
747 message to get a bit more predictability in the log output.
749 4. Added a call to search_tidyup() just before forking a reception process. In
750 theory, someone could use a lookup in the expansion of smtp_accept_max_
751 per_host which, without the tidyup, could leave open a database connection.
753 5. Added the variables $recipient_data and $sender_data which get set from a
754 lookup success in an ACL "recipients" or "senders" condition, or a router
755 "senders" option, similar to $domain_data and $local_part_data.
757 6. Moved the writing of debug_print from before to after the "senders" test
760 7. Change 4.31/66 (moving the time when the Received: is generated) caused
761 problems for message scanning, either using a data ACL, or using
762 local_scan() because the Received: header was not generated till after they
763 were called (in order to set the time as the time of reception completion).
764 I have revised the way this works. The header is now generated after the
765 body is received, but before the ACL or local_scan() are called. After they
766 are run, the timestamp in the header is updated.
772 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
773 before starting a queue runner without re-exec. This happened only when
774 deliver_drop_privilege was set or when the Exim user was set to root. The
775 effect of the bug was that timeouts during subsequent deliveries caused
776 crashes instead of being properly handled. The handler is now left at its
777 default (and expected) setting.
779 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
780 message, again when deliver_drop_privilege is set or Exim is run as root.
781 The bug described in (1) was not present in this case, but the tidying up
782 of the other signals was missing. I have made the two cases consistent.
784 3. The ignore_target_hosts setting on a manualroute router was being ignored
785 for hosts that were looked up using the /MX notation.
787 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
790 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
791 operated on the sender address. After changing the $sender_address to <>
792 for the sender address verify, Exim was re-instated it as the original
793 (before rewriting) address, but remembering that it had rewritten it, so it
794 wasn't rewriting it again. This bug also had the effect of breaking the
795 sender address verification caching when the sender address was rewritten.
797 6. The ignore_target_hosts option was being ignored by the ipliteral router.
798 This has been changed so that if the ip literal address matches
799 ignore_target_hosts, the router declines.
801 7. Added expansion conditions match_domain, match_address, and match_local_
802 part (NOT match_host).
804 8. The placeholder for the Received: header didn't have a length field set.
806 9. Added code to Exim itself and to exim_lock to test for a specific race
807 condition that could lead to file corruption when using MBX delivery. The
808 issue is with the lockfile that is created in /tmp. If this file is removed
809 after a process has opened it but before that process has acquired a lock,
810 there is the potential for a second process to recreate the file and also
811 acquire a lock. This could lead to two Exim processes writing to the file
812 at the same time. The added code performs the same test as UW imapd; it
813 checks after acquiring the lock that its file descriptor still refers to
816 10. The buffer for building added header lines was of fixed size, 8192 bytes.
817 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
820 11. Added the smtp_active_hostname option. If used, this will typically be made
821 to depend on the incoming interface address. Because $interface_address is
822 not set up until the daemon has forked a reception process, error responses
823 that can happen earlier (such as "too many connections") no longer contain
826 12. If an expansion in a condition on a "warn" statement fails because a lookup
827 defers, the "warn" statement is abandoned, and the next ACL statement is
828 processed. Previously this caused the whole ACL to be aborted.
830 13. Added the iplsearch lookup type.
832 14. Added ident_timeout as a log selector.
834 15. Added tls_certificate_verified as a log selector.
836 16. Added a global option tls_require_ciphers (compare the smtp transport
837 option of the same name). This controls incoming TLS connections.
839 17. I finally figured out how to make tls_require_ciphers do a similar thing
840 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
841 before starting the TLS session.
843 18. Tabs are now shown as \t in -bP output.
845 19. If the log selector return_path_on_delivery was set, Exim crashed when
846 bouncing a message because it had too many Received: header lines.
848 20. If two routers both had headers_remove settings, and the first one included
849 a superfluous trailing colon, the final name in the first list and the
850 first name in the second list were incorrectly joined into one item (with a
851 colon in the middle).
857 1. Added -C and -D options to the exinext utility, mainly to make it easier
858 to include in the automated testing, but these could be helpful when
859 multiple configurations are in use.
861 2. The exinext utility was not formatting the output nicely when there was
862 an alternate port involved in the retry record key, nor when there was a
863 message id as well (for retries that were specific to a specific message
864 and a specific host). It was also confused by IPv6 addresses, because of
865 the additional colons they contain. I have fixed the IPv4 problem, and
866 patched it up to do a reasonable job for IPv6.
868 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
869 delivery, the log line now contains "pipelined" if PIPELINING was used.
871 4. An SMTP transport process used to panic and die if the bind() call to set
872 an explicit outgoing interface failed. This has been changed; it is now
873 treated in the same way as a connect() failure.
875 5. A reference to $sender_host_name in the part of a conditional expansion
876 that was being skipped was still causing a DNS lookup. This no longer
879 6. The def: expansion condition was not recognizing references to header lines
880 that used bh_ and bheader_.
882 7. Added the _cache feature to named lists.
884 8. The code for checking quota_filecount in the appendfile transport was
885 allowing one more file than it should have been.
887 9. For compatibility with Sendmail, the command line option
895 and sets the incoming protocol and host name (for trusted callers). The
896 host name and its colon can be omitted when only the protocol is to be set.
897 Note the Exim already has two private options, -pd and -ps, that refer to
898 embedded Perl. It is therefore impossible to set a protocol value of "d" or
899 "s", but I don't think that's a major issue.
901 10. A number of refactoring changes to the code, none of which should affect
904 (a) The number of logging options was getting close to filling up the
905 32-bit word that was used as a bit map. I have split them into two classes:
906 those that are passed in the argument to log_write(), and those that are
907 only ever tested independently outside of that function. These are now in
908 separate 32-bit words, so there is plenty of room for expansion again.
909 There is no change in the user interface or the logging behaviour.
911 (b) When building, for example, log lines, the code previously used a
912 macro that called string_cat() twice, in order to add two strings. This is
913 not really sufficiently general. Furthermore, there was one instance where
914 it was actually wrong because one of the argument was used twice, and in
915 one call a function was used. (As it happened, calling the function twice
916 did not affect the overall behaviour.) The macro has been replaced by a
917 function that can join an arbitrary number of extra strings onto a growing
920 (c) The code for expansion conditions now uses a table and a binary chop
921 instead of a serial search (which was left over from when there were very
922 few conditions). Also, it now recognizes conditions like "pam" even when
923 the relevant support is not compiled in: a suitably worded error message is
924 given if an attempt is made to use such a condition.
926 11. Added ${time_interval:xxxxx}.
928 12. A bug was causing one of the ddress fields not to be passed back correctly
929 from remote delivery subprocesses. The field in question was not being
930 subsequently used, so this caused to problems in practice.
932 13. Added new log selectors queue_time and deliver_time.
934 14. Might have fixed a bug in maildirsizefile handling that threw up
935 "unexpected character" debug warnings, and recalculated the data
936 unnecessarily. In any case, I expanded the warning message to give more
939 15. Added the message "Restricted characters in address" to the statements in
940 the default ACL that block characters like @ and % in local parts.
942 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
943 Three changes have been made:
945 (a) There was a serious bug; a negative response to MAIL caused the whole
946 recipient domain to be cached as invalid, thereby blocking all messages
947 to all local parts at the same domain, from all senders. This bug has
948 been fixed. The domain is no longer cached after a negative response to
949 MAIL if the sender used is not empty.
951 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
954 (c) A new callout option, "use_sender" has been added for people who want
955 the modified behaviour.
961 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
964 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
965 indeed breaks things for older releases.
967 3. Added additional logging to the case where there is a problem reading data
968 from a filter that is running in a subprocess using a pipe, in order to
969 try to track down a specific problem.
971 4. Testing facility fudge: when running in the test harness and attempting
972 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
973 getting "No route to host". Convert this to a timeout.
975 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
978 6. Some OS don't have socklen_t but use size_t instead. This affects the
979 fifth argument of getsockopt() amongst other things. This is now
980 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
981 can be set for individual OS. I have set it for SunOS5, OSF1, and
982 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
983 some earlier ones do not.
985 7. Change 4.30/15 was not doing the test caselessly.
987 8. The standard form for an IPv6 address literal was being rejected by address
988 parsing in, for example, MAIL and RCPT commands. An example of this kind of
989 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
990 this, as well as the form without the "IPv6" on the front (but only when
991 address literals are enabled, of course).
993 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
995 10. Exim crashed if a message with an empty sender address specified by -f
996 encountered a router with an errors_to setting. This could be provoked only
1001 where an empty string was supplied; "<>" did not hit this bug.
1003 11. Installed PCRE release 4.5.
1005 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1006 remained set. It is now erased.
1008 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1009 times from message ids (which are base 36 rather than the normal 62).
1011 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1012 were being counted as actual protocol errors, and logged if the log
1013 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1014 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1015 and DATA following a set of rejected RCPTs do not count as protocol errors.
1016 In other words, Exim assumes they were pipelined, though this may not
1017 actually be the case. Of course, in all cases the client gets an
1018 appropriate error code.
1020 15. If a lookup fails in an ACL condition, a message about the failure may
1021 be available; it is used if testing the ACL cannot continue, because most
1022 such messages specify what the cause of the deferral is. However, some
1023 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1024 that caused an old message to be retained and used if a later statement
1025 caused a defer, replacing the real cause of the deferral.
1027 16. If an IP address had so many PTR records that the DNS lookup buffer
1028 was not large enough to hold them, Exim could crash while trying to process
1029 the truncated data. It now detects and logs this case.
1031 17. Further to 4.21/58, another change has been made: if (and only if) the
1032 first line of a message (the first header line) ends with CRLF, a bare LF
1033 in a subsequent header line has a space inserted after it, so as not to
1034 terminate the header.
1036 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1037 unnecessarily, used atoi() instead of strtol(), and didn't check the
1038 termination when getting file sizes from file names by regex.
1040 19. Completely re-implemented the support for maildirsize files, in the light
1041 of a number of problems with the previous contributed implementation
1042 (4.30/29). In particular:
1044 . If the quota is zero, the maildirsize file is maintained, but no quota is
1047 . If the maildir directory does not exist, it is created before any attempt
1048 to write a maildirsize file.
1050 . The quota value in the file is just a cache; if the quota is changed in
1051 the transport, the new value overrides.
1053 . A regular expression is available for excluding directories from the
1056 20. The autoreply transport checks the characters in options that define the
1057 message's headers; it allows continued headers, but it was checking with
1058 isspace() after an embedded newline instead of explicitly looking for a
1061 21. If all the "regular" hosts to which an address was routed had passed their
1062 expiry times, and had not reached their retry times, the address was
1063 bounced, even if fallback hosts were defined. Now Exim should go on to try
1066 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1067 equivalent code in the SMTP transport. Some hosts send humungous responses
1068 to HELO/EHLO, more than 1024 it seems.
1070 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1071 gives compiler warnings in some environments. I've now done it "properly",
1074 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1075 (because of problems with the built-in one) was declared to return uschar *
1076 instead of char *, causing compiler failure.
1078 25. Fixed a file descriptor leak when processing alias/forward files.
1080 26. Fixed a minor format string issue in dbfn.c.
1082 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1084 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1085 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1088 29. When a sender address is verified, it is cached, to save repeating the test
1089 when there is more than one recipient in a message. However, when the
1090 verification involves a callout, it is possible for different callout
1091 options to be set for different recipients. It is too complicated to keep
1092 track of this in the cache, so now Exim always runs a verification when a
1093 callout is required, relying on the callout cache for the optimization.
1094 The overhead is duplication of the address routing, but this should not be
1097 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
1098 to be verified with callout=postmaster, and the main callout worked but the
1099 postmaster check failed, the verification correctly failed. However, if a
1100 subsequent RCPT command asked for sender verification *without* the
1101 postmaster check, incorrect caching caused this verification also to fail,
1104 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1105 it was not caching the DNS options (qualify_single, search_parents) that
1106 were used when the lookup failed. A subsequent lookup with different
1107 options therefore always gave the same answer, though there were cases
1108 where it should not have. (Example: a "domains = !$mx_any" option on a
1109 dnslookup router: the "domains" option is always processed without any
1110 widening, but the router might have qualify_single set.) Now Exim uses the
1111 cached value only when the same options are set.
1113 32. Added John Jetmore's "exipick" utility to the distribution.
1115 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
1116 than a timeout (e.g. a certificate is required, and is not provided), an
1117 Exim server now closes the connection immediately. Previously it waited for
1118 the client to close - but if the client is SSL, it seems that they each
1119 wait for each other, leading to a delay before one of them times out.
1121 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1122 maintained 0.8.x compatibility because I don't think many are using it, and
1123 it is clearly obsolete.
1125 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1128 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1129 client certificate was expired. A simple patch fixes this, though I don't
1130 understand the full logic of why the verify callback is called multiple
1133 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1134 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1135 which causes problems with some clients (such as the Certicom SSL Plus
1136 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1137 disables the coutermeasure allowing Eudora to connect."
1139 38. Exim was not checking that a write() to a log file succeeded. This could
1140 lead to Bad Things if a log got too big, in particular if it hit a file
1141 size limit. Exim now panics and dies if it cannot write to a log file, just
1142 as it does if it cannot open a log file.
1144 39. Modified OS/Makefile-Linux so that it now contains
1146 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
1148 The two -D definitions ensure that Exim is compiled with large file
1149 support, which makes it possible to handle log files that are bigger than
1152 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
1153 instance) a domain was checked against a named list that involved a lookup,
1154 causing $domain_data to be set, then another domain was checked against the
1155 same list, then the first domain was re-checked, the value of $domain_data
1156 after the final check could be wrong. In particular, if the second check
1157 failed, it could be set empty. This bug probably also applied to
1160 41. The strip_trailing_dot option was not being applied to the address given
1161 with the -f command-line option.
1163 42. The code for reading a message's header from the spool was incrementing
1164 $received_count, but never initializing it. This meant that the value was
1165 incorrect (doubled) while delivering a message in the same process in which
1166 it was received. In the most common configuration of Exim, this never
1167 happens - a fresh exec is done - but it can happen when
1168 deliver_drop_privilege is set.
1170 43. When Exim logs an SMTP synchronization error - client data sent too soon -
1171 it now includes up to 150 characters of the unexpected data in the log
1174 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
1175 and building data strings. The size of both of these buffers was 10 000
1176 bytes - far larger than anybody would *ever* want, thought I. Needless to
1177 say, somebody hit the limit. I have increased the maximum line length to
1178 20 000 and the maximum data length of concatenated lines to 100 000. I have
1179 also fixed two bugs, because there was no checking on these buffers. Tsk,
1180 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
1181 buffer is too small.
1183 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
1184 lsearch lookups. Now it does.
1186 46. When parsing a route_list item in a manualroute router, a fixed-length
1187 buffer was used for the list of hosts. I made this 1024 bytes long,
1188 thinking that nobody would ever have a list of hosts that long. Wrong.
1189 Somebody had a whole pile of complicated expansion conditions, and the
1190 string was silently truncated, leading to an expansion error. It turns out
1191 that it is easier to change to an unlimited length (owing to other changes
1192 that have happened since this code was originally written) than to build
1193 structure for giving a limitation error. The length of the item that
1194 expands into the list of hosts is now unlimited.
1196 47. The lsearch lookup could not handle data where the length of text line was
1197 more than 4095 characters. Such lines were truncated, leading to shortened
1198 data being returned. It should now handle lines of any length.
1200 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
1201 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
1204 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
1205 replacing strings such as BIN_DIRECTORY in a source file: the replacement
1206 no longer happens in comment lines. A list of replacements is now placed
1207 at the head of all of the source files, except those whose only change is
1208 to replace PERL_COMMAND in the very first #! line.
1210 50. Replaced the slow insertion sort in queue.c, for sorting the list of
1211 messages on the queue, with a bottom-up merge sort, using code contributed
1212 by Michael Haardt. This should make operations like -bp somewhat faster on
1213 large queues. It won't affect queue runners, except when queue_run_in_order
1216 51. Installed eximstats 1.31 in the distribution.
1218 52. Added support for SRV lookups to the dnslookup router.
1220 53. If an ACL referred to $message_body or $message_body_end, the value was not
1221 reset for any messages that followed in the same SMTP session.
1223 54. The store-handling optimization for building very long strings was not
1224 differentiating between the different store pools. I don't think this
1225 actually made any difference in practice, but I've tidied it.
1227 55. While running the routers to verify a sender address, $sender_address
1228 was still set to the sender address. This is wrong, because when routing to
1229 send a bounce to the sender, it would be empty. Therefore, I have changed
1230 it so that, while verifying a sender address, $sender_address is set to <>.
1231 (There is no change to what happens when verifying a recipient address.)
1233 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
1234 target A or AAAA records (if not already returned) without resetting the
1235 qualify_single or search_parents options of the DNS resolver. These are
1236 inappropriate in this case because the targets of MX and SRV records must
1237 be FQDNs. A broken DNS record could cause trouble if it happened to have a
1238 target that, when qualified, matched something in the local domain. These
1239 two options are now turned off when doing these lookups.
1241 57. It seems that at least some releases of Reiserfs (which does not have the
1242 concept of a fixed number of inodes) returns zero and not -1 for the
1243 number of available inodes. This interacted badly with check_spool_inodes,
1244 which assumed that -1 was the "no such thing" setting. What I have done is
1245 to check that the total number of inodes is greater than zero before doing
1246 the test of how many are available.
1248 58. When a "warn" ACL statement has a log_message modifier, the message is
1249 remembered, and not repeated. This is to avoid a lot of repetition when a
1250 message has many recipients that cause the same warning to be written.
1251 Howewer, Exim was preserving the list of already written lines for an
1252 entire SMTP session, which doesn't seem right. The memory is now reset if a
1253 new message is started.
1255 59. The "rewrite" debugging flag was not showing the result of rewriting in the
1256 debugging output unless log_rewrite was also set.
1258 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
1259 of (int)(handle) when we know that handle contains (void *)(-1).
1261 61. The Exim daemon panic-logs an error return when it closes the incoming
1262 connection. However "connection reset by peer" seems to be common, and
1263 isn't really an error worthy of noting specially, so that particular error
1266 62. When Exim is trying to find all the local interfaces, it used to panic and
1267 die if the ioctl to get the interface flags failed. However, it seems that
1268 on at least one OS (Solaris 9) it is possible to have an interface that is
1269 included in the list of interfaces, but for which you get a failure error
1270 for this call. This happens when the interface is not "plumbed" into a
1271 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
1272 failure of the "get flags" call assumes that the interface is down.
1274 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
1275 makes life easier for people who are doing arithmetic on fields extracted
1276 from dates, where you often get leading zeros that should not be
1277 interpreted as octal.
1279 64. Added qualify_domain to the redirect router, to override the global
1282 65. If a pathologically long header line contained very many addresses (the
1283 report of this problem mentioned 10 000) and each of them was rewritten,
1284 Exim could use up a very large amount of memory. (It kept on making new
1285 copies of the header line as it rewrote, and never released the old ones.)
1286 At the expense of a bit more processing, the header rewriting function has
1287 been changed so that it no longer eats memory in this way.
1289 66. The generation of the Received: header has been moved from the time that a
1290 message starts to be received, to the time that it finishes. The timestamp
1291 in the Received: header should now be very close to that of the <= log
1292 line. There are two side-effects of this change:
1294 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
1295 logged header lines no longer include the local Received: line, because
1296 it has not yet been created. The same applies to a copy of the message
1297 that is returned to a non-SMTP sender when a message is rejected.
1299 (b) When a filter file is tested using -bf, no additional Received: header
1300 is added to the test message. After some thought, I decided that this
1303 This change does not affect the value of $received_for. It is still set
1304 after address rewriting, but before local_scan() is called.
1306 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
1308 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
1309 gave an unhelpful panic error message, and a defer error. I have managed to
1310 change this behaviour so that it now rejects any supplied certificate,
1311 which seems right, as the list of acceptable certificates is empty.
1313 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
1314 gave an unhelpful defer error. I have not managed to make this reject any
1315 supplied certificates, but the error message it gives is "no certificate
1316 supplied", which is not helpful.
1318 70. exigrep's output now also includes lines that are not associated with any
1319 message, but which match the given pattern. Implemented by a patch from
1320 Martin Sluka, which also tidied up the Perl a bit.
1322 71. Recipient callout verification, like sender verification, was using <> in
1323 the MAIL FROM command. This isn't really the right thing, since the actual
1324 sender may affect whether the remote host accepts the recipient or not. I
1325 have changed it to use the actual sender in the callout; this means that
1326 the cache record is now keyed on a recipient/sender pair, not just the
1327 recipient address. There doesn't seem to be a real danger of callout loops,
1328 since a callout by the remote host to check the sender would use <>.
1329 [SEE ABOVE: changed after hitting problems.]
1331 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
1332 temporary errors. However, in the case of such a code being given after
1333 the end of a data transmission (i.e. after ".") Exim was failing to write
1334 a retry record for the message. (Yes, there was some broken host that was
1335 actually sending 8xx at this point.)
1337 73. An unknown lookup type in a host list could cause Exim to panic-die when
1338 the list was checked. (An example that provoked this was putting <; in the
1339 middle of a list instead of at the start.) If this happened during a DATA
1340 ACL check, a -D file could be left lying around. This kind of configuration
1341 error no longer causes Exim to die; instead it causes a defer errror. The
1342 incident is still logged to the main and panic logs.
1344 74. Buglet left over from Exim 3 conversion. The message "too many messages
1345 in one connection" was written to the rejectlog but not the mainlog, except
1346 when address rewriting (yes!) was being logged.
1348 75. Added write_rejectlog option.
1350 76. When a system filter was run not as root (that is, when system_filter_user
1351 was set), the values of the $n variables were not being returned to the
1352 main process; thus, they were not subsequently available in the $sn
1355 77. Added +return_path_on_delivery log selector.
1357 78. A connection timeout was being treated differently from recipients deferred
1358 when testing hosts_max_try with a message that was older than the host's
1359 retry timeout. (The host should not be counted, thus allowing all hosts to
1360 be tried at least once before bouncing.) This may have been the cause of an
1361 occasionally reported bug whereby a message would remain on the queue
1362 longer than the retry timeout, but would be bounced if a delivery was
1363 forced. I say "may" because I never totally pinned down the problem;
1364 setting up timeout/retry tests is difficult. See also the next item.
1366 79. The ultimate address timeout was not being applied to errors that involved
1367 a combination of host plus message (for example, a timeout on a MAIL
1368 command). When an address resolved to a number of possible hosts, and they
1369 were not all tried for each delivery (e.g. because of hosts_max_try), a
1370 message could remain on the queue longer than the retry timeout.
1372 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
1375 81. Fixed an obscure SMTP outgoing bug which required at least the following
1376 conditions: (a) there was another message waiting for the same server;
1377 (b) the server returned 5xx to all RCPT commands in the first message so
1378 that the message was not completed; (c) the server dropped the connection
1379 or gave a negative response to the RSET that Exim sends to abort the
1380 transaction. The observed case was a dropped connection after DATA that had
1381 been sent in pipelining mode. That is, the server had advertised PIPELINING
1382 but was not implementing it correctly. The effect of the bug was incorrect
1383 behaviour, such as trying another host, and this could lead to a crash.
1389 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
1390 and daemon.c were passed as pointers to ints; they should have been
1391 pointers to socklen_t variables (which are typically unsigned ints).
1393 2. Some signed/unsigned type warnings in the os.c file for Linux have been
1396 3. Fixed a really odd bug that affected only the testing scheme; patching a
1397 certain fixed string in the binary changed the value of another string that
1398 happened to be identical to the end of the original first string.
1400 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
1401 name", it returns that address as the IP address. On some operating
1402 systems (e.g. Solaris), it also passes back the IP address string as the
1403 "host name". However, on others (e.g. Linux), it passes back an empty
1404 string. Exim wasn't checking for this, and was changing the host name to an
1405 empty string, assuming it had been canonicized.
1407 5. Although rare, it is permitted to have more than one PTR record for a given
1408 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
1409 all the names associated with an address, because they do in Solaris.
1410 However, it seems that they do not in Linux for data that comes from the
1411 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
1412 I found this out when I moved to a new Linux workstation and tried to run
1413 the Exim test suite.
1415 To get round this problem I have changed the code so that it now does its
1416 own call to the DNS to look up PTR records when searching for a host name.
1417 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
1418 addresses that are only in /etc/hosts are still found.
1420 This behaviour is, however, controlled by an option called host_lookup_
1421 order, which defaults to "bydns:byaddr". If people want to use the other
1422 order, or indeed, just use one or the other means of lookup, they can
1423 specify it in this variable.
1425 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
1426 some operating systems, this comes back from gethostbyaddr() as an empty
1427 string, and this is what Exim used to test for. However, it seems that in
1428 other systems, "." is yielded. Exim now tests for this case too.
1430 7. The values of check_spool_space and check_log_space are now held internally
1431 as a number of kilobytes instead of an absolute number of bytes. If a
1432 numbers is specified without 'K' or 'M', it is rounded up to the nearest
1433 kilobyte. This means that much larger values can be stored.
1435 8. Exim monitor: an attempt to get the action menu when not actually pointing
1436 at a message produces an empty menu entitled "No message selected". This
1437 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
1438 no entries in it ("Shell widget menu has zero width and/or height"). So I
1439 have added a single, blank menu entry in this case.
1441 9. Added ${quote_local_part.
1443 10. MIME decoding is now applied to the contents of Subject: header lines when
1446 11. Now that a reference to $sender_host_address automatically causes a reverse
1447 lookup to occur if necessary (4.13/18), there is no need to arrange for a
1448 host lookup before query-style lookups in lists that might use this
1449 variable. This has therefore been abolished, and the "net-" prefix is no
1450 longer necessary for query-style lookups.
1452 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
1453 have been a typo for LFLAGS, so it has been changed.
1455 13. The install script calls Exim with "-C /dev/null" in order to find the
1456 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
1457 to be output. Howeve, since Exim outputs its version number before the
1458 error, it didn't break the script. It just looked ugly. I fixed this by
1459 always allowing "-C /dev/null" if the caller is root.
1461 14. Ignore overlarge ACL variable number when reading spool file - insurance
1462 against a later release with more variables having written the file.
1464 15. The standard form for an IPv6 address literal was being rejected by EHLO.
1465 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1466 this, as well as the form without the "IPv6" on the front.
1468 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
1469 OS/Makefile-Darwin file.
1471 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
1472 only to LDAP libraries that do not have LDAP_OPT_DEREF.
1474 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
1475 variable. However, some OS use ints for inodes. I've added cast to long int
1476 to get rid of the compiler warning.
1478 19. I had forgotten to lock out "/../" in configuration file names when
1479 ALT_CONFIG_PREFIX was set.
1481 20. Routers used for verification do not need to specify transports. However,
1482 if such a router generated a host list, and callout was configured, Exim
1483 crashed, because it could not find a port number from the (non-existent)
1484 transport. It now assumes port 25 in this circumstance.
1486 21. Added the -t option to exigrep.
1488 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
1489 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
1490 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
1491 them is set, LOOKUP_LSEARCH is forced.
1493 23. "exim -bV" now outputs a list of lookups that are included in the binary.
1495 24. Added sender and host information to the "rejected by local_scan()" log
1496 line; previously there was no indication of these.
1498 25. Added .include_if_exists.
1500 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
1501 Linux. It turns out that not all file systems support this. Apparently some
1502 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
1503 do it.) To cope with this, the error EINVAL, which means that sync-ing is
1504 not supported on the file descriptor, is now ignored when Exim is trying to
1505 sync a directory. This applies only to Linux.
1507 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
1509 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
1510 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
1513 29. Added support for maildirsize files from supplied patch (modified a bit).
1515 30. The use of :fail: followed by an empty string could lead Exim to respond to
1516 sender verification failures with (e.g.):
1518 550 Verification failed for <xxx>
1519 550 Sender verify failed
1521 where the first response line was missing the '-' that indicates it is not
1522 the final line of the response.
1524 31. The loop for finding the name of the user that called Exim had a hardwired
1525 limit of 10; it now uses the value of finduser_retries, which is used for
1526 all other user lookups.
1528 32. Added $received_count variable, available in data and not_smtp ACLs, and at
1531 33. Exim was neglecting to zero errno before one call of strtol() when
1532 expanding a string and expecting an integer value. On some systems this
1533 resulted in spurious "integer overflow" errors. Also, it was casting the
1534 result into an int without checking.
1536 34. Testing for a connection timeout using "timeout_connect" in the retry rules
1537 did not work. The code looks as if it has *never* worked, though it appears
1538 to have been documented since at least releast 1.62. I have made it work.
1540 35. The "timeout_DNS" error in retry rules, also documented since at least
1541 1.62, also never worked. As it isn't clear exactly what this means, and
1542 clearly it isn't a major issue, I have abolished the feature by treating it
1543 as "timeout", and writing a warning to the main and panic logs.
1545 36. The display of retry rules for -brt wasn't always showing the error code
1548 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
1549 timeout_connect_A, timeout_connect_MX.
1551 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
1552 to the empty sender.
1554 39. The daemon was not analysing the content of -oX till after it had closed
1555 stderr and disconnected from the controlling terminal. This meant that any
1556 syntax errors were only noted on the panic log, and the return code from
1557 the command was 0. By re-arranging the code a little, I've made the
1558 decoding happen first, so such errors now appear on stderr, and the return
1559 code is 1. However, the actual setting up of the sockets still happens in
1560 the disconnected process, so errors there are still only recorded on the
1563 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
1564 connections (as happens on some IP stacks) was logged at start up time as
1565 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
1566 it from "IPv6 and IPv4", which means that two separate sockets are being
1569 41. The debug output for gethostbyname2() or getipnodebyname() failures now
1570 says whether AF_INET or AF_INET6 was passed as an argument.
1572 42. Exiwhat output was messed up when time zones were included in log
1575 43. Exiwhat now gives more information about the daemon's listening ports,
1576 and whether -tls-on-connect was used.
1578 44. The "port" option of the smtp transport is now expanded.
1580 45. A "message" modifier in a "warn" statement in a non-message ACL was being
1581 silently ignored. Now an error message is written to the main and panic
1584 46. There's a new ACL modifier called "logwrite" which writes to a log file
1585 as soon as it is encountered.
1587 47. Added $local_user_uid and $local_user_gid at routing time.
1589 48. Exim crashed when trying to verify a sender address that was being
1592 49. Exim was recognizing only a space character after ".include". It now also
1593 recognizes a tab character.
1595 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
1596 extracting the relevant information from the specification. The man page no
1597 longer contains scrambled data for the -d option, and I've added a section
1598 at the front about calling Exim under different names.
1600 51. Added "extra_headers" argument to the "mail" command in filter files.
1602 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
1605 53. Installed eximstats 1.29.
1607 54. Added transport_filter_timeout as a generic transport option.
1609 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
1610 Cc: header lines. This was required by RFC 822, but it not required by RFC
1613 56. Exim used to add From:, Date:, and Message-Id: header lines to any
1614 incoming messages that did not have them. Now it does so only if the
1615 message originates locally, that is, if there is no associated remote host
1616 address. When Resent- header lines are present, this applies to the Resent-
1617 lines rather than the non-Resent- lines.
1619 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
1620 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
1622 58. Messages for configuration errors now include the name of the main
1623 configuration file - useful now that there may be more than one file in a
1624 list (.included file names were always shown).
1626 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
1627 for those rare installations that do not start the daemon as root or run it
1628 setuid root. I've cut out the call to initgroups() if the daemon is not
1631 60. The Exim user and group can now be bound into the binary as text strings
1632 that are looked up at the start of Exim's processing.
1634 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
1636 62. Added $mailstore_basename variable.
1638 63. Installed patch to sieve.c from Michael Haardt.
1640 64. When Exim failed to open the panic log after failing to open the main log,
1641 the original message it was trying to log was written to stderr and debug
1642 output, but if they were not available (the usual case in production), it
1643 was lost. Now it is written to syslog before the two lines that record the
1644 failures to open the logs.
1646 65. Users' Exim filters run in subprocesses under the user's uid. It is
1647 possible for a "deliver" command or an alias in a "personal" command to
1648 provoke an address rewrite. If logging of address rewriting is configured,
1649 this fails because the process is not running as root or exim. There may be
1650 a better way of dealing with this, but for the moment (because 4.30 needs
1651 to be released), I have disabled address rewrite logging when running a
1652 filter in a non-root, non-exim process.
1658 1. The buildconfig auxiliary program wasn't quoting the value set for
1659 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
1660 not defined. This bug was masked in 4.22 by the effect that was fixed in
1663 2. Some messages that were rejected after a message id was allocated were
1664 shown as "incomplete" by exigrep. It no longer does this for messages that
1665 are rejected by local_scan() or the DATA or non-SMTP ACLs.
1667 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
1668 have allow_domain_literals set, the ID did not get logged in the <= line.
1669 Domain literals are now always recognized in Message-ID: header lines.
1671 4. The first argument for a ${extract expansion item is the key name or field
1672 number. Leading and trailing spaces in this item were not being ignored,
1673 causing some misleading effects.
1675 5. When deliver_drop_privilege was set, single queue runner processes started
1676 manually (i.e. by the command "exim -q") or by the daemon (which uses the
1677 same command in the process it spins off) were not dropping privilege.
1679 6. When the daemon running as "exim" started a queue runner, it always
1680 re-executed Exim in the spun-off process. This is a waste of effort when
1681 deliver_drop_privilege is set. The new process now just calls the
1682 queue-runner function directly.
1688 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
1691 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
1692 ignored. Though the use of -oP was forcing the writing of a pid file, it
1693 was always written to the default place.
1695 3. If the message "no IP address found for host xxxx" is generated during
1696 incoming verification, it is now followed by identification of the incoming
1697 connection (so you can more easily find what provoked it).
1699 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
1701 5. Added some features to "harden" Exim a bit more against certain attacks:
1703 (a) There is now a build-time option called FIXED_NEVER_USERS that can
1704 be put in Local/Makefile. This is like the never_users runtime option,
1705 but it cannot be overridden. The default setting is "root".
1707 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
1708 prefix string with which any file named in a -C command line option
1711 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
1712 is retained for -C and -D only if the caller of Exim is root. Without
1713 it, the exim user may also use -C and -D and retain privilege.
1715 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
1716 command line option is disabled.
1718 6. Macro names set by the -D option must start with an upper case letter, just
1719 like macro names defined in the configuration file.
1721 7. Added "dereference=" facility to LDAP.
1723 8. Two instances of the typo "uknown" in the source files are fixed.
1725 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
1726 the Configure-Makefile script screwed up while processing it.
1728 10. Incorporated PCRE 4.4.
1730 11. The SMTP synchronization check was not operating right at the start of an
1731 SMTP session. For example, it could not catch a HELO sent before the client
1732 waited for the greeting. There is now a check for outstanding input at the
1733 point when the greeting is written. Because of the duplex, asynchronous
1734 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
1735 way, but not yet received, when the check is performed.
1737 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
1738 on TCP/IP sockets, because this apparently causes some broken clients to
1741 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
1742 unchanged) from the Cygwin maintainer.
1744 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
1745 was supported instead of testing for mbx. Effectively a typo.
1747 15. The spa authenticator server code was not checking that the input it
1748 received was valid base64.
1750 16. The debug output line for the "set" modifier in ACLs was not showing the
1751 name of the variable that was being set.
1753 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
1755 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
1756 lookup on its own. Something else was needed to trigger the lookup. For
1757 example, a match in host_lookup or the need for a host name in a host list.
1758 Now, if $sender_host_name is referenced and the host name has not yet been
1759 looked up, a lookup is performed. If the lookup fails, the variable remains
1760 empty, and $host_lookup_failed is set to "1".
1762 19. Added "eqi" as a case-independent comparison operator.
1764 20. The saslauthd authentication condition could segfault if neither service
1765 nor realm was specified.
1767 21. If an overflowing value such as "2048M" was set for message_size_limit, the
1768 error message that was logged was misleading, and incoming SMTP
1769 connections were dropped. The message is now more accurate, and temporary
1770 errors are given to SMTP connections.
1772 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
1773 (except RSET) with a 421 error, until QUIT is received. However, it was
1774 failing to send a response to QUIT.
1776 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
1777 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
1778 ACL is now run after the helo_try_verify_hosts code.
1780 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
1781 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
1782 case-independent, so other case variants are also recognized). Apparently
1783 some systems use these upper case variants.
1785 25. If more than two messages were waiting for the same host, and a transport
1786 filter was specified for the transport, Exim sent two messages over the
1787 same TCP/IP connection, and then failed with "socket operation on non-
1788 socket" when it tried to send the third.
1790 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
1792 27. The extern definition of crypt16() in expand.c was not being excluded when
1793 the OS had its own crypt16() function.
1795 28. Added bounce_return_body as a new option, and bounce_return_size_limit
1796 as a preferred synonym for return_size_limit, both as an option and as an
1799 29. Added LIBS=-liconv to OS/Makefile-OSF1.
1801 30. Changed the default configuration ACL to relax the local part checking rule
1802 for addresses that are not in any local domains. For these addresses,
1803 slashes and pipe symbols are allowed within local parts, but the sequence
1804 /../ is explicitly forbidden.
1806 31. SPA server authentication was not clearing the challenge buffer before
1809 32. log_message in a "warn" ACL statement was writing to the reject log as
1810 well as to the main log, which contradicts the documentation and doesn't
1811 seem right (because no rejection is happening). So I have stopped it.
1813 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
1814 However, I am unable to do any testing of this.
1816 34. Fixed an infelicity in the appendfile transport. When checking directories
1817 for a mailbox, to see if any needed to be created, it was accidentally
1818 using path names with one or more superfluous leading slashes; tracing
1819 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
1821 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
1822 contents are added to the log line that is written for every discarded
1823 recipient. (Previously a log_message setting was ignored.)
1825 36. The ${quote: operator now quotes the string if it is empty.
1827 37. The install script runs exim in order to find its version number. If for
1828 some reason other than non-existence or emptiness, which it checks, it
1829 could not run './exim', it was installing it with an empty version number,
1830 i.e. as "exim-". This error state is now caught, and the installation is
1833 38. An argument was missing from the function that creates an error message
1834 when Exim fails to connect to the socket for saslauthd authentication.
1835 This could cause Exim to crash, or give a corrupted message.
1837 39. Added isip, isip4, and isip6 to ${if conditions.
1839 40. The ACL variables $acl_xx are now saved with the message, and can be
1840 accessed later in routers, transports, and filters.
1842 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
1843 strings in the file are not string-expanded.
1845 42. If a MAIL command specified a SIZE value that was too large to fit into an
1846 int variable, the check against message_size_limit failed. Such values are
1847 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
1848 day this will have to be increased, but I don't think I want to be around
1849 when emails are that large.
1856 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
1857 iconv() is not standard in FreeBSD.
1859 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
1860 IPv6 enabled. The observed symptom was a segmentation fault on return from
1861 the function os_common_find_running_interfaces() in src/os.c.
1863 3. In the check_special_case() function in daemon.c I had used "errno" as an
1864 argument name, which causes warnings on some systems. This was basically a
1865 typo, since it was named "eno" in the comments!
1867 4. The code that waits for the clock to tick (at a resolution of some fraction
1868 of a second) so as to ensure message-id uniqueness was always waiting for
1869 at least one whole tick, when it could have waited for less. [This is
1870 almost certainly not relevant at current processor speeds, where it is
1871 unlikely to ever wait at all. But we try to future-proof.]
1873 5. The function that sleeps for a time interval that includes fractions of a
1874 second contained a race. It did not block SIGALRM between setting the
1875 timer, and suspending (a couple of lines later). If the interval was short
1876 and the sigsuspend() was delayed until after it had expired, the suspension
1877 never ended. On busy systems this could lead to processes getting stuck for
1880 6. Some uncommon configurations may cause a lookup to happen in a queue runner
1881 process, before it forks any delivery processes. The open lookup caching
1882 mechanism meant that the open file or database connection was passed into
1883 the delivery process. The problem was that delivery processes always tidy
1884 up cached lookup data. This could cause a problem for the next delivery
1885 process started by the queue runner, because the external queue runner
1886 process does not know about the closure. So the next delivery process
1887 still has data in the lookup cache. In the case of a file lookup, there was
1888 no problem because closing a file descriptor in a subprocess doesn't affect
1889 the parent. However, if the lookup was caching a connection to a database,
1890 the connection was closed, and the second delivery process was likely to
1891 see errors such as "PGSQL: query failed: server closed the connection
1892 unexpectedly". The problem has been fixed by closing all cached lookups
1893 in a queue runner before running a delivery process.
1895 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
1896 seem to have the "const" qualifier which it has on other OS. I've
1899 8. Change 4.21/2 was too strict. It is only if there are two authenticators
1900 *of the same type* (client or server) with the same public name that an
1901 error should be diagnosed.
1903 9. When Exim looked up a host name for an IP address, but failed to find the
1904 original IP address when looking up the host name (a safety check), it
1905 output the message "<ip address> does not match any IP for NULL", which was
1906 confusing, to say the least. The bug was that the host name should have
1907 appeared instead of "NULL".
1909 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
1910 user that is built into the binary, and the -C or -D options is used, root
1911 privilege is dropped before the configuration file is read. In addition,
1912 logging is switched to stderr instead of the normal log files. If the
1913 configuration then re-defines the Exim user, the unprivileged environment
1914 is probably not what is expected, so Exim logs a panic warning message (but
1917 However, if deliver_drop_privilege is set, the unprivileged state may well
1918 be exactly what is intended, so the warning has been cut out in that case,
1919 and Exim is allowed to try to write to its normal log files.
1925 1. smtp_return_error_details was not giving details for temporary sender
1926 or receiver verification errors.
1928 2. Diagnose a configuration error if two authenticators have the same public
1931 3. Exim used not to create the message log file for a message until the first
1932 delivery attempt. This could be confusing when incoming messages were held
1933 for policy or load reasons. The message log file is now created at the time
1934 the message is received, and an initial "Received" line is written to it.
1936 4. The automatically generated man page for command line options had a minor
1937 bug that caused no ill effects; however, a more serious problem was that
1938 the procedure for building the man page automatically didn't always
1939 operate. Consequently, release 4.20 contains an out-of-date version. This
1940 shouldn't happen again.
1942 5. When building Exim with embedded Perl support, the script that builds the
1943 Makefile was calling 'perl' to find its compile-time parameters, ignoring
1944 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
1946 6. The freeze_tell option was not being used for messages that were frozen on
1947 arrival, either by an ACL or by local_scan().
1949 7. Added the smtp_incomplete_transaction log selector.
1951 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
1952 was accepting AUTH without a new EHLO.
1954 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
1955 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
1958 10. Logging of TCP/IP connections (when configured) now happens in the main
1959 daemon process instead of the child process, so that the TCP/IP connection
1960 count is more accurate (but it can never be perfect).
1962 11. The use of "drop" in a nested ACL was not being handled correctly in the
1963 outer ACL. Now, if condition failure induced by the nested "drop" causes
1964 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
1965 or "require"), the connection is dropped.
1967 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
1968 that yield "discard" can now be used with an "accept" or a "discard" verb,
1969 but an error is generated for any others (because I can't see a useful way
1970 to define what should happen).
1972 13. When an ACL is read dynamically from a file (or anywhere else), the lines
1973 are now processed in the same way as lines in the Exim configuration file.
1974 In particular, continuation lines are supported.
1976 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
1978 15. Added -ti meaning -t -i.
1980 16. Check for letters, digits, hyphens, and dots in the names of dnslist
1981 domains, and warn by logging if others are found.
1983 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
1984 returned from GIFCONF when Exim is trying to find the list of interfaces on
1985 a host. The code in os.c has been modified to copy each ifreq to an aligned
1986 structure in all cases.
1988 Also, in some cases, the returned ifreq's were being copied to a 'struct
1989 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
1990 means the last couple of bytes of an IPv6 address could be chopped if the
1991 ifreq contained only a normal sockaddr (14 bytes storage).
1993 18. Named domain lists were not supported in the hosts_treat_as_local option.
1994 An entry such as +xxxx was not recognized, and was treated as a literal
1997 19. Ensure that header lines added by a DATA ACL are included in the reject log
1998 if the ACL subsequently rejects the message.
2000 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2001 MD5 (which is deprecated).
2003 21. When testing a filter file using -bf, Exim was writing a message when it
2004 took the sender from a "From " line in the message, but it was not doing so
2005 when it took $return_path from a Return-Path: header line. It now does.
2007 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2008 with a valid header line field name (a series of printing characters
2009 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2011 23. Changed "disc" in the source to "disk" to conform to the documentation and
2012 the book and for uniformity.
2014 24. Ignore Sendmail's -Ooption=value command line item.
2016 25. When execve() failed while trying to run a command in a pipe transport,
2017 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2018 could be confused with a return value of 69 from the command itself. This
2019 has been changed to 127, the value the shell returns if it is asked to run
2020 a non-existent command. The wording for the related log line suggests a
2021 non-existent command as the problem.
2023 26. If received_header_text expands to an empty string, do not add a Received:
2024 header line to the message. (Well, it adds a token one on the spool, but
2025 marks it "old" so that it doesn't get used or transmitted.)
2027 27. Installed eximstats 1.28 (addition of -nt option).
2029 28. There was no check for failure on the call to getsockname() in the daemon
2030 code. This can fail if there is a shortage of resources on the system, with
2031 ENOMEM, for example. A temporary error is now given on failure.
2033 29. Contrary to the C standard, it seems that in some environments, the
2034 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2035 program. Exim now does this explicitly; it affects the formatting of
2036 timestamps using strftime().
2038 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2039 complaints. I've now initialized all the variables, to avoid this.
2041 32. Header lines added by a system filter were not being "seen" during
2042 transport-time rewrites.
2044 33. The info_callback() function passed to OpenSSL is set up with type void
2045 (*)(SSL *, int, int), as described somewhere. However, when calling the
2046 function (actually a macro) that sets it up, the type void(*)() is
2047 expected. I've put in a cast to prevent warnings from picky compilers.
2049 34. If a DNS black list lookup found a CNAME record, but there were no A
2050 records associated with the domain it pointed at, Exim crashed.
2052 35. If a DNS black list lookup returned more than one A record, Exim ignored
2053 all but the first. It now scans all returned addresses if a particular IP
2054 value is being sought. In this situation, the contents of the
2055 $dnslist_value variable are a list of all the addresses, separated by a
2058 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2059 to accept a host name and all its aliases if the forward lookup for any of
2060 them yielded the IP address of the incoming connection. Now it accepts only
2061 those names whose forward lookup yields the correct IP address. Any other
2062 names are discarded. This closes a loophole whereby a rogue DNS
2063 administrator could create reverse DNS records to break through a
2064 wildcarded host restriction in an ACL.
2066 37. If a user filter or a system filter that ran in a subprocess used any of
2067 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2068 the wrong values were passed to the pipe command ($thisaddress had the
2069 value of $0, $0 had the value of $1, etc). This bug was introduced by
2070 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2072 38. Improved the line breaking for long SMTP error messages from ACLs.
2073 Previously, if there was no break point between 40 and 75 characters, Exim
2074 left the rest of the message alone. Two changes have been made: (a) I've
2075 reduced the minimum length to 35 characters; (b) if it can't find a break
2076 point between 35 and 75 characters, it looks ahead and uses the first one
2077 that it finds. This may give the occasional overlong line, but at least the
2078 remaining text gets split now.
2080 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2081 file descriptors might be low, and that setting 1000 would always raise it.
2082 It turns out that in some environments, the limit is already over 1000 and
2083 that lowering it causes trouble. So now Exim takes care not to decrease it.
2085 40. When delivering a message, the value of $return_path is set to $sender_
2086 address at the start of routing (routers may change the value). By an
2087 oversight, this default was not being set up when an address was tested by
2088 -bt or -bv, which affected the outcome if any router or filter referred to
2091 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2092 log only when "message" or "log_message" are set. However, if one of the
2093 conditions was an address verification, or a call to a nested ACL, the
2094 messages generated by the underlying test were being passed through. This
2095 no longer happens. The underlying message is available in $acl_verify_
2096 message for both "message" and "log_message" expansions, so it can be
2097 passed through if needed.
2099 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2100 new expansion $bh_ to give the encoded byte string without charset
2101 translation. Translation happens only if iconv() is available; HAVE_ICONV
2102 indicates this at build time. HEADERS_CHARSET gives the charset to
2103 translate to; headers_charset can change it in the configuration, and
2104 "headers charset" can change it in an individual filter file.
2106 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
2107 that creates RFC 2047 encoded "words" labels them as that charset instead
2108 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2109 expansion operator; (ii) when Exim creates a From: line for a local
2110 message; (iii) when a header line is rewritten to include a "phrase" part.
2112 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2113 buggy, causing it to skip the first lines of messages whose message ID
2114 ended in 'D'. This would not have bitten before Exim release 4.14, because
2115 message IDs were unlikely to end in 'D' before then. The effect was to have
2116 incorrect size information for certain domains.
2118 45. #include "config.h" was missing at the start of the crypt16.c module. This
2119 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2122 46. If there was a timeout during a "random" callout check, Exim treated it as
2123 a failure of the random address, and carried on sending RSET and the real
2124 address. If the delay was just some slowness somewhere, the response to the
2125 original RCPT would be taken as a response to RSET and so on, causing
2126 mayhem of various kinds.
2128 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2129 when I implemented it. It didn't allow for the fact that some option values
2130 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2131 the right test for positive values.
2133 48. Domain names in DNS records are case-independent. Exim always looks them up
2134 in lower case. Some resolvers return domain names in exactly the case they
2135 appear in the zone file, that is, they may contain uppercase letters. Not
2136 all resolvers do this - some return always lower case. Exim was treating a
2137 change of case by a resolver as a change of domain, similar to a widening
2138 of a domain abbreviation. This triggered its re-routing code and so it was
2139 trying to route what was effectively the same domain again. This normally
2140 caused routing to fail (because the router wouldn't handle the domain
2141 twice). Now Exim checks for this case specially, and just changes the
2142 casing of the domain that it ultimately uses when it transmits the message
2145 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
2148 50. If a filter generated a file delivery with a non-absolute name (possible if
2149 no home directory exists for the router), the forbid_file option was not
2152 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
2153 the existing equality matching.
2155 52. Exim was using ints instead of ino_t variables in some places where it was
2156 dealing with inode numbers.
2158 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
2159 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
2160 called TMPDIR, and if it finds it is different, it changes its value.
2162 54. The smtp_printf() function is now made available to local_scan() so
2163 additional output lines can be written before returning. There is also an
2164 smtp_fflush() function to enable the detection of a dropped connection.
2165 The variables smtp_input and smtp_batched_input are exported to
2168 55. Changed the default runtime configuration: the message "Unknown user"
2169 has been removed from the ACL, and instead placed on the localuser router,
2170 using the cannot_route_message feature. This means that any verification
2171 failures that generate their own messages won't get overridden. Similarly,
2172 the "Unrouteable address" message that was in the ACL for unverifiable
2173 relay addresses has also been removed.
2175 56. Added hosts_avoid_esmtp to the smtp transport.
2177 57. The exicyclog script was not checking for the esoteric option
2178 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
2179 will work only if exicyclog is run under the appropriate euid.
2181 58. Following a discussion on the list, the rules by which Exim recognises line
2182 endings on incoming messages have been changed. The -dropcr and drop_cr
2183 options are now no-ops, retained only for backwards compatibility. The
2184 following line terminators are recognized: LF CRLF CR. However, special
2185 processing applies to CR:
2187 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
2188 nor a local message in the state where . is a terminator.
2190 (ii) If a bare CR is encountered in a header line, an extra space is added
2191 after the line terminator so as not to end the header. The reasoning
2192 behind this is that bare CRs in header lines are most likely either
2193 to be mistakes, or people trying to play silly games.
2195 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
2196 was being incorrectly given as 18 bytes larger than it should have been.
2197 This is a VOB (very old bug).
2199 60. This may never have affected anything current, but just in case it has:
2200 When the local host is found other than at the start of a list of hosts,
2201 the local host, those with the same MX, and any that follow, are discarded.
2202 When the list in question was part of a longer list of hosts, the following
2203 hosts (not currently being processed) were also being discarded. This no
2204 longer happens. I'm not sure if this situation could ever has previously
2207 61. Added the "/MX" feature to lists of hosts in the manualroute and query
2210 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
2211 header. This is something that is recommended in a new Internet Draft, and
2212 is something that is documented as being done by Sendmail. There are two
2213 possible values. For messages generated by the autoreply transport, Exim
2216 Auto-Submitted: auto-replied
2218 whereas for all other generated messages (e.g. bounces) it adds
2220 Auto-Submitted: auto-generated
2222 63. The "personal" condition in filters now includes a test for the
2223 Auto-Submitted: header. If it contains the string "auto-" the message it
2224 not considered personal.
2226 64. Added rcpt_include_affixes as a generic transport option.
2228 65. Added queue_only_override (default true).
2230 66. Added the syslog_duplication option.
2232 67. If what should have been the first header line of a message consisted of
2233 a space followed by a colon, Exim was mis-interpreting it as a header line.
2234 It isn't of course - it is syntactically invalid and should therefore be
2235 treated as the start of the message body. The misbehaviour could have
2236 caused a number of strange effects, including loss of data in subsequent
2237 header lines, and spool format errors.
2239 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
2240 client host had authenticated. This control can now be exercised by an ACL
2241 for more flexibility.
2243 69. By default, callouts do not happen when testing with -bh. There is now a
2244 variant, -bhc, which does actually run the callout code, including
2245 consulting and updating the callout cache.
2247 70. Added support for saslauthd authentication, courtesy of Alexander
2250 71. If statvfs() failed on the spool or log directories while checking their
2251 size for availability, Exim confusingly gave the error "space shortage".
2252 Furthermore, in debugging mode it crashed with a floating point exception.
2253 These checks are done if check_{spool,log}_{space,inodes} are set, and when
2254 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
2255 serious problem, Exim now writes to the main and panic logs when this
2256 happens, with details of the failure. It then refuses to accept the
2257 incoming message, giving the message "spool directory problem" or "log
2258 directory problem" with a 421 code for SMTP messages.
2260 72. When Exim is about to re-exec itself, it ensures that the file descriptors
2261 0, 1, and 2 exist, because some OS complain for execs without them (see
2262 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
2263 descriptors. However, the code omitted to check that the open succeeded,
2264 causing mysterious errors if for some reason the permissions on /dev/null
2265 got screwed. Now Exim writes a message to the main and panic logs, and
2266 bombs out if it can't open /dev/null.
2268 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
2269 interact so that it is all more flexible. It is supposed to remain
2270 backwards compatible. Also added extra_local_interfaces.
2272 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
2273 to bomb out with an assertion failure - to the client this appears as a
2274 connection drop. This problem occurs in the part of the code that was taken
2275 from the Samba project. Fortunately, the assertion is in a very simple
2276 function, so I have fixed this by reproducing the function inline in the
2277 one place where it is called, and arranging for authentication to fail
2278 instead of killing the process with assert().
2280 75. The SPA client code was not working when the server requested OEM rather
2281 than Unicode encoding.
2283 76. Added code to make require_files with a specific uid setting more usable in
2284 the case where statting the file as root fails - usually a non-root-mounted
2285 NFS file system. When this happens and the failure is EACCES, Exim now
2286 forks a subprocess and does the per-uid checking as the relevant uid.
2288 77. Added process_log_path.
2290 78. If log_file_path was not explicitly set, a setting of check_log_space or
2291 check_log_inodes was ignored.
2293 79. If a space check for the spool or log partitions fails, the incident is now
2294 logged. Of course, in the latter case the data may get lost...
2296 80. Added the %p formatting code to string_format() so that it can be used to
2297 print addresses in debug_print(). Adjusted all the address printing in the
2298 debugging in store.c to use %p rather than %d.
2300 81. There was a concern that a line of code in smtp_in.c could overflow a
2301 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
2302 initially expressed, the concern was not well-founded, because trailing
2303 spaces are removed early. However, if the trailing spaces were followed by
2304 a NULL, they did not get removed, so the overflow was possible. Two fixes
2307 (a) I re-wrote the offending code in a cleaner fashion.
2308 (b) If an incoming SMTP command contains a NULL character, it is rejected
2311 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
2312 runs initgroups(), so that if the Exim user is in any additional groups,
2313 they will be used during message reception.
2319 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob