Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
[exim.git] / doc / doc-txt / ChangeLog
3e46c1aa 1$Cambridge: exim/doc/doc-txt/ChangeLog,v 1.259 2005/11/14 15:28:48 ph10 Exp $
3Change log file for Exim from version 4.21
6Exim version 4.60
9PH/01 Two changes to the default runtime configuration:
11 (1) Move the checks for relay_from_hosts and authenticated clients from
12 after to before the (commented out) DNS black list checks.
14 (2) Add control=submission to the relay_from_hosts and authenticated
15 clients checks, on the grounds that messages accepted by these
16 statements are most likely to be submissions.
5de37277 17
18PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
20 (1) Generate an error if the third argument for the ${prvs expansion is
21 not a single digit.
23 (2) Treat a missing third argument of ${prvscheck as if it were an empty
24 string.
26 (3) Reset the variables that are obtained from the first argument of
27 ${prvscheck and used in the second argument before leaving the code,
28 because their memory is reclaimed, so using them afterwards may do
29 silly things.
31 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
32 one (it's much easier than Tom thought :-).
34 (5) Because of (4), we can now allow for the use of $prvscheck_result
35 inside the third argument.
cb9328de 36
37PH/03 For some reason, the default setting of PATH when running a command from
38 a pipe transport was just "/usr/bin". I have changed it to
39 "/bin:/usr/bin".
42 anything to be listed in the output from -bV.
b2f5a032 43
44PH/05 When a filter generated an autoreply, the entire To: header line was
45 quoted in the delivery log line, like this:
47 => >A.N.Other <ano@some.domain> <original@ddress> ...
49 This has been changed so that it extracts the operative address. There
50 may be more than one such address. If so, they are comma-separated, like
51 this:
53 => >ano@some.domain,ona@other.domain <original@ddress> ...
55PH/06 When a client host used a correct literal IP address in a HELO or EHLO
56 command, (for example, EHLO []) and the client's IP address was
57 not being looked up in the rDNS to get a host name, Exim was showing the
58 IP address twice in Received: lines, even though the IP addresses were
59 identical. For example:
61 Received: from [] (helo=[])
63 However, if the real host name was known, it was omitting the HELO data
64 if it matched the actual IP address. This has been tidied up so that it
65 doesn't show the same IP address twice.
67PH/07 When both +timestamp and +memory debugging was on, the value given by
68 $tod_xxx expansions could be wrong, because the tod_stamp() function was
69 called by the debug printing, thereby overwriting the timestamp buffer.
70 Debugging no longer uses the tod_stamp() function when +timestamp is set.
72PH/08 When the original message was included in an autoreply transport, it
73 always said "this is a copy of the message, including all the headers",
74 even if body_only or headers_only was set. It now gives an appropriate
75 message.
77PH/09 Applied a patch from the Sieve maintainer which:
79 o fixes some comments
80 o adds the (disabled) notify extension core
81 o adds some debug output for the result of if/elsif tests
82 o points to the current vacation draft in the documentation
83 and documents the missing references header update
85 and most important:
87 o fixes a bug in processing the envelope test (when testing
88 multiple envelope elements, the last element determinted the
89 result)
91PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
92 Electronic Mail") by including:
94 Auto-submitted: auto-generated
96 in the messages that it generates (bounce messages and others, such as
97 warnings). In the case of bounce messages for non-SMTP mesages, there was
98 also a typo: it was using "Auto_submitted" (underscore instead of
99 hyphen). Since every message generated by Exim is necessarily in response
100 to another message, thes have all been changed to:
102 Auto-Submitted: auto-replied
104 in accordance with these statements in the RFC:
106 The auto-replied keyword:
108 - SHOULD be used on messages sent in direct response to another
109 message by an automatic process,
111 - MUST NOT be used on manually-generated messages,
113 - MAY be used on Delivery Status Notifications (DSNs) and Message
114 Disposition Notifications (MDNs),
116 - MUST NOT be used on messages generated by automatic or periodic
117 processes, except for messages which are automatic responses to
118 other messages.
120PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
121 to the default Received: header definition.
456682f5 122
5de37277 123
124Exim version 4.54
127PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
128 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
129 It now does.
131PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
132 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
134PH/03 Typo: missing ".o" in src/pcre/Makefile.
136PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
137 header line, restrict the check to what is listed in RFCs 2369 and 2929.
138 Also, for "Auto-Submitted", treat anything other than "no" as
139 non-personal, in accordance with RFC 3834. (Previously it treated
140 anything starting "auto-" as non-personal.)
142TF/01 The control=submission/name=... option had a problem with syntax
143 errors if the name included a slash character. The /name= option
144 now slurps the rest of the string, so it can include any characters
145 but it must come last in the list of options (after /sender_retain
146 or /domain=).
148PH/05 Some modifications to the interface to the fake nameserver for the new
149 testing suite.
151PH/06 Add ${if def:sender_address {(envelope-from <$sender_address>)\n\t}}
152 into the default Received: header string.
9a799bc0 155
156Exim version 4.53
159TK/01 Added the "success_on_redirect" address verification option. See
160 NewStuff for rationale and an example.
162PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
164PH/02 Patch to exigrep to allow it to work on syslog lines.
166PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
167 fread() to skip over the body file's header line, because in Cygwin the
168 header line is locked and is inaccessible.
170PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
171 co-exist for some time) to make it clear that it is the Exim ID that is
172 referenced, not the Message-ID: header line.
174PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
175 string_format() function, because snprintf() does not exist on all
176 operating systems.
178PH/06 The use of forbid_filter_existstest now also locks out the use of the
179 ${stat: expansion item.
181PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
182 protocol synchronization error", to keep the pedants happy.
184PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
185 well as for IRIX systems, when gcc is being used. See the host.c source
186 file for comments.
188PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
190PH/10 Named domain lists were not working if used in a queue_smtp_domains
191 setting.
193PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
194 transport and to the smtp transport in LMTP mode.
196TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
198PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
199 run a filter in a subprocess. This could lead to confusion in subsequent
200 lookups in the parent process. There should also be a search_tidyup() at
201 the end of the subprocess.
203PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
204 only if the host matched helo_try_verify_hosts, which caused the
205 verification to occur when the EHLO/HELO command was issued. The ACL just
206 tested the remembered result. Now, if a previous verification attempt has
207 not happened, "verify = helo" does it there and then.
209JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
211TK/03 Fix log output including CR from clamd.
213PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
214 exist provoked a memory error which could cause a segfault.
216PH/15 Installed PCRE 6.2
218PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
220PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
221 of the problem. Specifically, suggested +O2 rather than +O1 for the
222 HP-UX compiler.
224PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
226PH/20 If a delivery was routed to a non-standard port by means of an SRV
227 record, the port was not correctly logged when the outgoing_port log
228 selector was set (it logged the transort's default port).
230PH/21 Added support for host-specific ports to manualroute, queryprogram,
231 fallback_hosts, and "hosts" in the smtp transport.
233PH/22 If the log selector "outgoing_port" is set, the port is now also given on
234 host errors such as "Connection refused".
236PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
237 authentication with radiusclient 0.4.9:
239 - Error returned from rc_read_config was caught wrongly
240 - Username/password not passed on to radius server due to wrong length.
242 The presumption is that some radiusclient API changes for 4.51/PH/17
243 were not taken care of correctly. The code is still untested by me (my
244 Linux distribution still has 0.3.2 of radiusclient), but it was
245 contributed by a Radius user.
247PH/24 When doing a callout, the value of $domain wasn't set correctly when
248 expanding the "port" option of the smtp transport.
250TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
251 while reading a MIME header. Thanks to Tom Hughes for a patch.
253PH/24 Include config.h inside local_scan.h so that configuration settings are
254 available.
256PH/25 Make $smtp_command_argument available after all SMTP commands. This means
257 that in an ACL for RCPT (for example), you can examine exactly what was
258 received.
260PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
261 commands, but it was not correctly comparing the address with the actual
262 client host address. Thus, it would show the EHLO address in Received:
263 header lines when this was not necessary.
265PH/27 Added the % operator to ${eval:}.
267PH/28 Exim tries to create and chdir to its spool directory when it starts;
268 it should be ignoring failures (because with -C, for example, it has lost
269 privilege). It wasn't ignoring creation failures other than "already
270 exists".
272PH/29 Added "crypteq" to the list of supported features that Exim outputs when
273 -bV or -d is used.
275PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
276 because an input line was too long, either on its own, or by virtue of
277 too many continuations, the temporary file was not being removed, and the
278 return code was incorrect.
aa2b5c79 279
280PH/31 Missing "BOOL" in function definition in filtertest.c.
282PH/32 Applied Sieve patches from the maintainer.
284TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
286PH/33 Added "verify = not_blind".
288PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
289 Local/Makefile (with some defaults set). These are used in built scripts
290 such as exicyclog, but they have never been used in the exim_install
291 script (though there are many overriding facilities there). I have
292 arranged that the exim_install script now takes note of these two
293 settings.
295PH/35 Installed configuration files for Dragonfly.
297PH/36 When a locally submitted message by a trusted user did not contain a
298 From: header, and the sender address was obtained from -f or from an SMTP
299 MAIL command, and the trusted user did not use -F to supply a sender
300 name, $originator_name was incorrectly used when constructing a From:
301 header. Furthermore, $originator_name was used for submission mode
302 messages from external hosts without From: headers in a similar way,
303 which is clearly wrong.
305PH/37 Added control=suppress_local_fixups.
307PH/38 When log_selector = +received_sender was set, and the addition of the
308 sender made the log line's construction buffer exactly full, or one byte
309 less than full, an overflow happened when the terminating "\n" was
310 subsequently added.
312PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
313 when the result of a list match is failure because a DNS lookup failed.
315PH/40 RM_COMMAND is now used in the building process.
317PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
318 the "build-* directories that it finds.
320PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
321 address in a domain literal was a prefix of an interface address.
323PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
324 when verifying a sender address, unless rewrite_headers is false.
326PH/44 Wrote a long comment about why errors_to addresses are verified as
327 recipients, not senders.
329TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
330 the ratelimit ACL was added.
332PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
334PH/46 Added extra information about PostgreSQL errors to the error string.
336PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
337 avoiding the need to install special zones in a real server. This is
338 backwards compatible; if it can't find the fake resolver, it drops back.
339 Thus, both old and new test suites can be run.
341TF/02 Added util/
343TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
344 clock is set back in time.
346TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
347 Candler <>.
349TF/05 The fix for PH/43 was not completely correct; widen_domains is always
350 OK for addresses that are the result of redirections.
352PH/48 A number of further additions for the benefit of the new test suite,
353 including a fake gethostbyname() that interfaces to the fake DNS resolver
354 (see PH/47 above).
356TF/06 The fix for widen_domains has also been applied to qualify_single and
357 search_parents which are the other dnslookup options that can cause
358 header rewrites.
360PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
361 type ("H").
363PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
365TF/07 Exim produced the error message "an SRV record indicated no SMTP
366 service" if it encountered an MX record with an empty target hostname.
367 The message is now "an MX or SRV record indicated no SMTP service".
369TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
370 if the DNS of the sending site is misconfigured. This is quite a
371 common situation. This change restores the behaviour of treating a
372 helo verification defer as a failure.
374PH/51 If self=fail was set on a router, the bounce message did not include the
375 actual error message.
bbe902f0 377
378Exim version 4.52
381TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
383PH/01 When a transport filter timed out in a pipe delivery, and the pipe
384 command itself ended in error, the underlying message about the transport
385 filter timeout was being overwritten with the pipe command error. Now the
386 underlying error message should be appended to the second error message.
388TK/01 Fix poll() being unavailable on Mac OSX 10.2.
390PH/02 Reduce the amount of output that "make" produces by default. Full output
391 can still be requested.
393PH/03 The warning log line about a condition test deferring for a "warn" verb
394 was being output only once per connection, rather than after each
395 occurrence (because it was using the same function as for successful
396 "warn" verbs). This seems wrong, so I have changed it.
398TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
399 it should not have, which might have caused a crash in the right
400 circumstances, but probably never did.
402PH/04 Installed a modified version of Tony Finch's patch to make submission
403 mode fix the return path as well as the Sender: header line, and to
404 add a /name= option so that you can make the user's friendly name appear
405 in the header line.
407TF/03 Added the control = fakedefer ACL modifier.
409TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
410 Mark Lowes for thorough testing.
870f6ba8 411
412TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
414TK/03 Merged latest SRS patch from Miles Wilton.
416PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
417 with the definition in sysexits.h (which is #included earlier).
418 Fortunately, Exim does not actually use EX_OK. The code used to try to
419 preserve the sysexits.h value, by assumimg that macro definitions were
420 scanned for macro replacements. I have been disabused of this notion,
421 so now the code just undefines EX_OK before #including unistd.h.
11d337a4 422
423PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
424 in the smtp transport. When a block could not be written in a single
425 write() function, the timeout was being re-applied to each part-write.
426 This seems wrong - if the receiver was accepting one byte at a time it
427 would take for ever. The timeout is now adjusted when this happens. It
428 doesn't have to be particularly precise.
430TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
431 details. Thanks to Chris Webb <> for the patch!
433PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
434 without a domain if the check to <postmaster@domain> fails.
436SC/01 Eximstats: added -xls and the ability to specify output files
437 (patch written by Frank Heydlauf).
439SC/02 Eximstats: use FileHandles for outputing results.
441SC/03 Eximstats: allow any combination of xls, txt, and html output.
443SC/04 Eximstats: fixed display of large numbers with -nvr option
445SC/05 Eximstats: fixed merging of reports with empty tables.
447SC/06 Eximstats: added the -include_original_destination flag
449SC/07 Eximstats: removed tabs and trailing whitespace.
451TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
453TK/06 MBOX spool code: Add real "From " MBOX separator line
454 so the .eml file is really in mbox format (even though
455 most programs do not really care). Patch from Alex Miller.
457TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
458 The latter is generated from $received_to and is only set if the
459 message has one envelope recipient. SA can use these headers,
460 obviously out-of-the-box. Patch from Alex Miller.
462PH/08 The ${def test on a variable was returning false if the variable's
463 value was "0", contrary to what the specification has always said!
464 The result should be true unless the variable is empty.
466PH/09 The syntax error of a character other than { following "${if
467 def:variable_name" (after optional whitespace) was not being diagnosed.
468 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
469 accidental colon was present, for example, could give incorrect results.
471PH/10 Tidied the code in a number of places where the st_size field of a stat()
472 result is used (not including appendfile, where other changes are about
473 to be made).
475PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
476 This involved changing a lot of size variables from int to off_t. It
477 should work with maildirs and everything.
479TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
480 spamd dying while we are connected to it.
482TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
483 <>
485PH/12 Applied Alex Kiernan's patch for the API change for the error callback
486 function for BDB 4.3.
488PH/13 Changed auto_thaw such that it does not apply to bounce messages.
490PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
491 the sources for PCRE have been re-arranged and more files are now
492 involved.
494PH/15 The code I had for printing potentially long long variables in PH/11
495 above was not the best (it lost precision). The length of off_t variables
496 is now inspected at build time, and an appropriate printing format (%ld
497 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
498 to be "long long int" or "long int". This is needed for the internal
499 formatting function string_vformat().
b1c749bb 500
501PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
502 the configuration file to be ":syslog", then the script "guesses" where
503 the logs files are, rather than using the compiled in default. In our
504 case the guess is not the same as the compiled default, so the script
505 suddenly stopped working when I started to use syslog. The patch checks
506 to see if log_file_path is "". If so, it attempts to read it from exim
507 with no configuration file to get the compiled in version, before it
508 falls back to the previous guessing code."
510TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
511 implementing BATV in an Exim configuration. See NewStuff for the gory
512 details.
514PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
515 Makefile that are specific to HP-UX.
517PH/18 If the "use_postmaster" option was set for a recipient callout together
518 with the "random" option, the postmaster address was used as the MAIL
519 FROM address for the random test, but not for the subsequent recipient
520 test. It is now used for both.
522PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
523 patch removes a few documentation additions to RFC 3028, because the
524 latest draft now contains them. It adds the new en;ascii-case comparator
525 and a new error check for 8bit text in MIME parts. Comparator and
526 require names are now matched exactly. I enabled the subaddress
527 extension, but it is not well tested yet (read: it works for me)."
529PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
530 rework some of the code of TK/09 above to avoid the hardwired use of
531 "%lld" and "long long". Replaced the call to snprintf() with a call to
532 string_vformat().
534PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
535 records point to non-existent hosts", "retry timeout exceeded", and
536 "retry time not reached for any host after a long failure period".
ca02eafb 537
538PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
539 experimental DomainKeys support:
541 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
542 (2) On an error such as an illegally used "control", the wrong name for
543 the control was given.
545 These problems did NOT occur unless DomainKeys support was compiled.
547PH/23 Added daemon_startup_retries and daemon_startup_sleep.
549PH/24 Added ${if match_ip condition.
551PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
552 databases so that it will be absolutely obvious if a crash occurs in the
553 DB library. This is a regular occurrence (often caused by mis-matched
554 db.h files).
ff790e47 556PH/26 Insert a lot of missing (void) casts for functions such as chown(),
557 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
558 picked up on a user's system that detects such things. There doesn't seem
559 to be a gcc warning option for this - only an attribute that has to be
560 put on the function's prototype. It seems that in Fedora Core 4 they have
561 set this on a number of new functions. No doubt there will be more in due
562 course.
ff790e47 563
564PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
565 specify a transport. However, if an address that was verified by such a
566 router was the subject of a callout, Exim crashed because it tried to
567 read the rcpt_include_affixes from the non-existent transport. Now it
568 just assumes that the setting of that option is false. This bug was
569 introduced by 4.51/PH/31.
571PH/28 Changed -d+all to exclude +memory, because that information is very
572 rarely of interest, but it makes the output a lot bigger. People tend to
573 do -d+all out of habit.
575PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
576 code in os-type was giving problems when lives in lib64, like on
577 x86_64 Fedora Core.
579PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
580 aren't the modern standard, and it seems that some systems' include files
581 don't always have them. Exim was already checking for some of the newer
582 ones like T_AAAA, and defining it itself. I've added checks for all the
583 record types that Exim uses.
585PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
586 not automatically generating a new one, as it is supposed to. This
587 prevented TLS from working. If the file did exist, but contained invalid
588 data, a new version was generated, as expected. It was only the case of a
589 non-existent file that was broken.
591TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
592 with a change in libdomainkeys > 0.64.
594TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
595 from DNS. If the selector record carries the flag, it now has
596 precedence over the domain-wide flag.
598TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
600PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
601 the use of a port name that isn't defined in /etc/services) occurred, the
602 message was deferred as in a normal delivery, and thus remained on the
603 spool, instead of being failed because of the mua_wrapper setting. This
604 is now fixed, and I tidied up some of the mua_wrapper messages at the
605 same time.
607SC/08 Eximstats: whilst parsing the mainlog(s), store information about
608 the messages in a hash of arrays rather than using individual hashes.
609 This is a bit cleaner and results in dramatic memory savings, albeit
610 at a slight CPU cost.
612SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
613 as requested by Marc Sherman.
615SC/10 Eximstats: added histograms for user specified patterns as requested
616 by Marc Sherman.
618SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
620PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
621 fopen() in the content-scanning modules that did not already have it.
e7ad8a65 623
624Exim version 4.51
627TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
628 doc/experimental-spec.txt for details. (
2f079f46 630TK/02 Fix ACL "control" statement not being available in MIME ACL.
632TK/03 Fix ACL "regex" condition not being available in MIME ACL.
634PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
635 to test Sieve filters that use "vacation".
637PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
638 that changes the way the GnuTLS parameters are stored in the cache file.
639 The new format can be generated externally. For backward compatibility,
640 if the data in the cache doesn't make sense, Exim assumes it has read an
641 old-format file, and it generates new data and writes a new file. This
642 means that you can't go back to an older release without removing the
643 file.
645PH/03 A redirect router that has both "unseen" and "one_time" set does not
646 work if there are any delivery delays because "one_time" forces the
647 parent to be marked "delivered", so its unseen clone is never tried
648 again. For this reason, Exim now forbids the simultaneous setting of
649 these two options.
651PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
652 redirected to themselves ("homonym" addresses). Read the long ChangeLog
653 entry if you want to know the details. The fix, however, neglected to
654 consider the case when local delivery batching is involved. The test for
655 "previously delivered" was not happening when checking to see if an
656 address could be batched with a previous (undelivered) one; under
657 certain circumstances this could lead to multiple deliveries to the same
c2c19e9d 658 address.
660PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
661 in its include files, and this causes problems building Exim.
663PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
664 header_syntax) but Exim was just ignoring anything given after a slash.
665 In particular, this caused confusion with an attempt to use "verify =
666 reverse_host_lookup/defer_ok". An error is now given when options are
667 supplied for verify items that do not have them. (Maybe reverse_host_
668 lookup should have a defer_ok option, but that's a different point.)
670PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
671 defined by RFC 821) to 2048, because there were problems with some AUTH
672 commands, and RFC 1869 says the size should be increased for extended
673 SMTP commands that take arguments.
675PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
676 Finch).
678PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
679 "unknown" error; now it says that the functionality isn't in the binary.
8d67ada3 680
681PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
682 an address' error message when a string expansion fails (syntax or
683 whatever). Otherwise the password may appear in the log. Following change
684 PH/42 below, there is no longer a chance of it appearing in a bounce
685 message.
49c2d5ea 686
687PH/11 Installed exipick version 20050225.0 from John Jetmore.
689PH/12 If the last host in a fallback_hosts list was multihomed, only the first
690 of its addresses was ever tried. (Bugzilla bug #2.)
692PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
693 the result incorrectly in the debug output. (It correctly added a newline
694 to what was transported.)
696TF/01 Added $received_time.
698PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
699 commented out examples of how to interface to a virus scanner and to
700 SpamAssassin. Also added commented examples of av_scanner and
701 spamd_address settings.
703PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
704 and controls are allowed in which ACLs. There were a couple of minor
705 errors. Some of the entries in the conditions table (which is a table of
706 where they are NOT allowed) were getting very unwieldy; rewrote them as a
707 negation of where the condition IS allowed.
709PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
711PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
712 header file does not have a version number, so I've had to invent a new
713 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
714 API. The code is untested by me (my Linux distribution still has 0.3.2 of
715 radiusclient), but it was contributed by a Radius user.
717PH/18 Installed Lars Mainka's patch for the support of CRL collections in
718 files or directories, for OpenSSL.
720PH/19 When an Exim process that is running as root has to create an Exim log
721 file, it does so in a subprocess that runs as exim:exim so as to get the
722 ownership right at creation (otherwise, other Exim processes might see
723 the file with the wrong ownership). There was no test for failure of this
724 fork() call, which would lead to the process getting stuck as it waited
725 for a non-existent subprocess. Forks do occasionally fail when resources
726 run out. I reviewed all the other calls to fork(); they all seem to check
727 for failure.
729PH/20 When checking for unexpected SMTP input at connect time (before writing
730 the banner), Exim was not dealing correctly with a non-positive return
731 from the read() function. If the client had disconnected by this time,
732 the result was a log entry for a synchronization error with an empty
733 string after "input=" when read() returned zero. If read() returned -1
734 (an event I could not check), uninitialized data bytes were printed.
735 There were reports of junk text (parts of files, etc) appearing after
736 "input=".
738PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
740PH/22 Added support for macro redefinition, and (re)definition in between
741 driver and ACL definitions.
743PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
744 forgetting to use the resulting value; it was using the unexpanded value.
746PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
747 hadn't been configured. The fix is from Juergen Kreileder, who
748 understands it better than I do:
750 "Here's what I see happening with three configured cyrus_sasl
751 authenticators configured (plain, login, cram-md5):
753 On startup auth_cyrus_sasl_init() gets called for each of these.
754 This means three calls to sasl_listmech() without a specified mech_list.
755 => SASL tests which mechs of all available mechs actually work
756 => three warnings about OTP not working
757 => the returned list contains: plain, login, cram-md5, digest-md5, ...
759 With the patch, sasl_listmech() also gets called three times. But now
760 SASL's mech_list option is set to the server_mech specified in the the
761 authenticator. Or in other words, the answer from sasl_listmech()
762 gets limited to just the mech you're testing for (which is different
763 for each call.)
764 => the return list contains just 'plain' or 'login', 'cram-md5' or
765 nothing depending on the value of ob->server_mech.
767 I've just tested the patch: Authentication still works fine,
768 unavailable mechs specified in the exim configuration are still
769 caught, and the auth.log warnings about OTP are gone."
771PH/25 When debugging is enabled, the contents of the command line are added
772 to the debugging output, even when log_selector=+arguments is not
773 specified.
775PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
776 answer is "GNU", and only if the return is "GNU/something" is the answer
777 "Linux".
779PH/27 $acl_verify_message is now set immediately after the failure of a
780 verification in an ACL, and so is available in subsequent modifiers. In
781 particular, the message can be preserved by coding like this:
783 warn !verify = sender
784 set acl_m0 = $acl_verify_message
786 Previously, $acl_verify_message was set only while expanding "message"
787 and "log_message" when a very denied access.
789PH/28 Modified OS/os.c-Linux with
791 -#ifndef OS_LOAD_AVERAGE
792 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
794 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
795 nomenclature these days.)
797PH/29 Installed patch from the Sieve maintainer that adds the options
798 sieve_useraddress and sieve_subaddress to the redirect router.
800PH/30 In these circumstances:
801 . Two addresses routed to the same list of hosts;
802 . First host does not offer TLS;
803 . First host accepts first address;
804 . First host gives temporary error to second address;
805 . Second host offers TLS and a TLS session is established;
806 . Second host accepts second address.
807 Exim incorrectly logged both deliveries with the TLS parameters (cipher
808 and peerdn, if requested) that were in fact used only for the second
809 address.
7e8bec7a 810
811PH/31 When doing a callout as part of verifying an address, Exim was not paying
812 attention to any local part prefix or suffix that was matched by the
813 router that accepted the address. It now behaves in the same way as it
814 does for delivery: the affixes are removed from the local part unless
815 rcpt_include_affixes is set on the transport.
817PH/32 Add the sender address, as F=<...>, to the log line when logging a
818 timeout during the DATA phase of an incoming message.
820PH/33 Sieve envelope tests were broken for match types other than :is. I have
821 applied a patch sanctioned by the Sieve maintainer.
c688b954 822
823PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
824 the uid or gid is negative. A case of a negative gid caused this to be
825 noticed. The fix allows for either to be negative.
827PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
828 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
829 overlooked.
831PH/36 The change PH/12 above was broken. Fixed it.
833PH/37 Exim used to check for duplicate addresses in the middle of routing, on
834 the grounds that routing the same address twice would always produce the
835 same answer. This might have been true once, but it is certainly no
836 longer true now. Routing a child address may depend on the previous
837 routing that produced that child. Some complicated redirection strategies
838 went wrong when messages had multiple recipients, and made Exim's
839 behaviour dependent on the order in which the addresses were given.
841 I have moved the duplicate checking until after the routing is complete.
842 Exim scans the addresses that are assigned to local and remote
843 transports, and removes any duplicates. This means that more work will be
844 done, as duplicates will always all be routed, but duplicates are
845 presumably rare, so I don't expect this is of any significance.
847 For deliveries to pipes, files, and autoreplies, the duplicate checking
848 still happens during the routing process, since they are not going to be
849 routed further.
851PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
852 It corrects a timeout issue with spamd. This is Ian's comment: "The
853 background is that sometimes spamd either never reads data from a
854 connection it has accepted, or it never writes response data. The exiscan
855 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
856 blindly assumes that writes won't block so it may never time out."
858PH/39 Allow G after quota size as well as K and M.
860PH/40 The value set for $authenticated_id in an authenticator may not contain
861 binary zeroes or newlines because the value is written to log lines and
862 to spool files. There was no check on this. Now the value is run through
863 the string_printing() function so that such characters are converted to
864 printable escape sequences.
866PH/41 $message_linecount is a new variable that contains the total number of
867 lines in the message. Compare $body_linecount, which is the count for the
868 body only.
870PH/42 Exim no longer gives details of delivery errors for specific addresses in
871 bounce and delay warning messages, except in certain special cases, which
872 are as follows:
874 (a) An SMTP error message from a remote host;
875 (b) A message specified in a :fail: redirection;
876 (c) A message specified in a "fail" command in a system filter;
877 (d) A message specified in a FAIL return from the queryprogram router;
878 (e) A message specified by the cannot_route_message router option.
880 In these cases only, Exim does include the error details in bounce and
881 warning messages. There are also a few cases where bland messages such
882 as "unrouteable address" or "local delivery error" are given.
884PH/43 $value is now also set for the "else" part of a ${run expansion.
886PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
887 being worked on, but at least Exim now implements the latest version to
888 play with."
890PH/45 In a pipe transport, although a timeout while waiting for the pipe
891 process to complete was treated as a delivery failure, a timeout while
892 writing the message to the pipe was logged, but erroneously treated as a
893 successful delivery. Such timeouts include transport filter timeouts. For
894 consistency with the overall process timeout, these timeouts are now
895 treated as errors, giving rise to delivery failures by default. However,
896 there is now a new Boolean option for the pipe transport called
897 timeout_defer, which, if set TRUE, converts the failures into defers for
898 both kinds of timeout. A transport filter timeout is now identified in
899 the log output.
901PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
902 systems where "make" and "gmake" are different, calling "gmake" at top
903 level broke things. I've arranged for the value of $(MAKE) to be passed
904 from the Makefile to this script so that it can call the same version of
905 "make".
7982096b 907
908A note about Exim versions 4.44 and 4.50
911Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
912changes of various kinds. As a consequence, a big documentation update was
913needed. This delayed the release for rather longer than seemed good, especially
914in the light of a couple of (minor) security issues. Therefore, the changes
915that fixed bugs were backported into 4.43, to create a 4.44 maintenance
916release. So 4.44 and 4.50 are in effect two different branches that both start
917from 4.43.
919I have left the 4.50 change log unchanged; it contains all the changes since
9204.43. The change log for 4.44 is below; many of its items are identical to
921those for 4.50. This seems to be the most sensible way to preserve the
922historical information.
f7b63901 925Exim version 4.50
928 1. Minor wording change to the doc/README.SIEVE file.
139059f6 930 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5fe762f6 931 computation of the current number of files was incorrect.
495ae4b0 932
933 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
934 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
935 place.
937 4. Give more explanation in the error message when the command for a transport
938 filter fails to execute.
940 5. There are several places where Exim runs a non-Exim command in a
941 subprocess. The SIGUSR1 signal should be disabled for these processes. This
942 was being done only for the command run by the queryprogram router. It is
943 now done for all such subprocesses. The other cases are: ${run, transport
944 filters, and the commands run by the lmtp and pipe transports.
946 6. Added CONFIGURE_GROUP build-time option.
948 7. Some older OS have a limit of 256 on the maximum number of file
949 descriptors. Exim was using setrlimit() to set 1000 as a large value
950 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
951 systems. I've change it so that if it can't get 1000, it tries for 256.
35edf2ff 952
953 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
954 was an oversight, and furthermore, ever since the addition of extra
955 controls (e.g. 4.43/32), the checks on when to allow different forms of
956 "control" were broken. There should now be diagnostics for all cases when a
957 control that does not make sense is encountered.
959 9. Added the /retain_sender option to "control=submission".
96110. $recipients is now available in the predata ACL (oversight).
96311. Tidy the search cache before the fork to do a delivery from a message
964 received from the command line. Otherwise the child will trigger a lookup
965 failure and thereby defer the delivery if it tries to use (for example) a
966 cached ldap connection that the parent has called unbind on.
96812. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
969 of $address_data from the recipient verification was clobbered by the
970 sender verification.
97213. The value of address_data from a sender verification is now available in
973 $sender_address_data in subsequent conditions in the ACL statement.
97514. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
97715. Added a new option "connect=<time>" to callout options, to set a different
978 connection timeout.
98016. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
981 was its contents. (It was OK if the option was not defined at all.)
98317. A "Completed" log line is now written for messages that are removed from
984 the spool by the -Mrm option.
98618. New variables $sender_verify_failure and $recipient_verify_failure contain
987 information about exactly what failed.
98919. Added -dd to debug only the daemon process.
99120. Incorporated Michael Haardt's patch to ldap.c for improving the way it
992 handles timeouts, both on the server side and network timeouts. Renamed the
993 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
99521. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
996 It is now set to "smtps".
99822. $host_address is now set to the target address during the checking of
999 ignore_target_hosts.
100123. When checking ignore_target_hosts for an ipliteral router, no host name was
1002 being passed; this would have caused $sender_host_name to have been used if
1003 matching the list had actually called for a host name (not very likely,
1004 since this list is usually IP addresses). A host name is now passed as
1005 "[x.x.x.x]".
100724. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1008 code that specifies a non-restarting handler (typically sigaction() in
1009 modern systems) in an attempt to fix a rare and obscure crash bug.
101125. Narrowed the window for a race in the daemon that could cause it to ignore
1012 SIGCHLD signals. This is not a major problem, because they are used only to
1013 wake it up if nothing else does.
101526. A malformed maildirsize file could cause Exim to calculate negative values
1016 for the mailbox size or file count. Odd effects could occur as a result.
1017 The maildirsize information is now recalculated if the size or filecount
1018 end up negative.
102027. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1021 support for a long time. Removed HAVE_SYS_VFS_H.
102328. Installed the latest version of exipick from John Jetmore.
102529. In an address list, if the pattern was not a regular expression, an empty
1026 subject address (from a bounce message) matched only if the pattern was an
1027 empty string. Non-empty patterns were not even tested. This was the wrong
1028 because it is perfectly reasonable to use an empty address as part of a
1029 database query. An empty address is now tested by patterns that are
1030 lookups. However, all the other forms of pattern expect the subject to
1031 contain a local part and a domain, and therefore, for them, an empty
1032 address still always fails if the pattern is not itself empty.
103430. Exim went into a mad DNS loop when attempting to do a callout where the
1035 host was specified on an smtp transport, and looking it up yielded more
1036 than one IP address.
103831. Re-factored the code for checking spool and log partition space into a
1039 function that finds that data and another that does the check. The former
1040 is then used to implement four new variables: $spool_space, $log_space,
1041 $spool_inodes, and $log_inodes.
104332. The RFC2047 encoding function was originally intended for short strings
1044 such as real names; it was not keeping to the 75-character limit for
1045 encoded words that the RFC imposes. It now respects the limit, and
1046 generates multiple encoded words if necessary. To be on the safe side, I
1047 have increased the buffer size for the ${rfc2047: expansion operator from
1048 1024 to 2048 bytes.
105033. It is now permitted to omit both strings after an "if" condition; if the
1051 condition is true, the result is "true". As before, when the second string
1052 is omitted, a false condition yields an empty string. This makes it less
1053 cumbersome to write custom ACL and router conditions.
105534. Failure to deliver a bounce message always caused it to be frozen, even if
1056 there was an errors_to setting on the router. The errors_to setting is now
1057 respected.
105935. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1060 canonical form (fully expanded) before being placed in
1061 $sender_host_address.
106336. The table in the code that translates DNS record types into text (T_A to
1064 "A" for instance) was missing entries for NS and CNAME. It is just possible
1065 that this could have caused confusion if both these types were looked up
1066 for the same domain, because the text type is used as part of Exim's
1067 per-process caching. But the chance of anyone hitting this buglet seems
1068 very small.
107037. The dnsdb lookup has been extended in a number of ways.
1072 (1) There is a new type, "zns", which walks up the domain tree until it
1073 finds some nameserver records. It should be used with care.
1075 (2) There is a new type, "mxh", which is like "mx" except that it returns
1076 just the host names, not the priorities.
1078 (3) It is now possible to give a list of domains (or IP addresses) to be
1079 looked up. The behaviour when one of the lookups defers can be
1080 controlled by a keyword.
7bb56e1f 1081
ea3bc19b 1082 (4) It is now possible to specify the separator character for use when
7bb56e1f 1083 multiple records are returned.
33397d19 1084
108538. The dnslists ACL condition has been extended: it is now possible to supply
1086 a list of IP addresses and/or domains to be looked up in a particular DNS
1087 domain.
108939. Added log_selector=+queue_time_overall.
109140. When running the queue in the test harness, wait just a tad after forking a
1092 delivery process, to get repeatability of debugging output.
109441. Include certificate and key file names in error message when GnuTLS fails
1095 to set them up, because the GnuTLS error message doesn't include the name
1096 of the failing file when there is a problem reading it.
109842. Allow both -bf and -bF in the same test run.
110043. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
110244. The "Exiscan patch" is now merged into the mainline Exim source.
110445. Sometimes the final signoff response after QUIT could fail to get
1105 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1106 < 0 before doing a fflush(). This bug looks as though it goes back to the
1107 introduction of TLS in release 3.20, but "sometimes" must have been rare
1108 because the tests only now provoked it.
111046. Reset the locale to "C" after calling embedded Perl, in case it was changed
1111 (this can affect the format of dates).
111347. exim_tidydb, when checking for the continued existence of a message for
1114 which it has found a message-specific retry record, was not finding
1115 messages that were in split spool directories. Consequently, it was
1116 deleting retry records that should have stayed in existence.
111848. Steve fixed some bugs in eximstats.
112049. The SPA authentication driver was not abandoning authentication and moving
1121 on to the next authenticator when an expansion was forced to fail,
1122 contradicting the general specification for all authenticators. Instead it
1123 was generating a temporary error. It now behaves as specified.
112550. The default ordering of permitted cipher suites for GnuTLS was pessimal
1126 (the order specifies the preference for clients). The order is now AES256,
1127 AES128, 3DES, ARCFOUR128.
112951. Small patch to Sieve code - explicitly set From: when generating an
1130 autoreply.
113252. Exim crashed if a remote delivery caused a very long error message to be
1133 recorded - for instance if somebody sent an entire SpamAssassin report back
1134 as a large number of 550 error lines. This bug was coincidentally fixed by
1135 increasing the size of one of Exim's internal buffers (big_buffer) that
1136 happened as part of the Exiscan merge. However, to be on the safe side, I
1137 have made the code more robust (and fixed the comments that describe what
1138 is going on).
114053. Now that there can be additional text after "Completed" in log lines (if
1141 the queue_time_overall log selector is set), a one-byte patch to exigrep
1142 was needed to allow it to recognize "Completed" as not the last thing in
1143 the line.
114554. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1146 patch that reportedly fixes this has been added. I am not expert enough to
1147 create a test for it. This is what the patch creator wrote:
1149 "I found a little strange behaviour of ldap code when working with
1150 Windows 2003 AD Domain, where users was placed in more than one
1151 Organization Units. When I tried to give exim partial DN, the exit code
1152 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1153 But simultaneously result of request was absolutely normal ldap result,
1154 so I produce this patch..."
1156 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1157 REFERENCE, so I have modified the code to exclude the patch when that macro
1158 is not defined.
116055. Some experimental protocols are using DNS PTR records for new purposes. The
1161 keys for these records are domain names, not reversed IP addresses. The
1162 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1163 leaves it alone. Component reversal etc. now happens only for IP addresses.
ea3a6f44 1164 CAN-2005-0021
7102e136 1165
116656. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
116857. Double the size of the debug message buffer (to 2048) so that more of very
1169 long debug lines gets shown.
117158. The exicyclog utility now does better if the number of log files to keep
1172 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
117459. Two changes related to the smtp_active_hostname option:
1176 (1) $smtp_active_hostname is now available as a variable.
1177 (2) The default for smtp_banner uses $smtp_active_hostname instead
1178 of $primary_hostname.
118060. The host_aton() function is supposed to be passed a string that is known
1181 to be a valid IP address. However, in the case of IPv6 addresses, it was
1182 not checking this. This is a hostage to fortune. Exim now panics and dies
1183 if the condition is not met. A case was found where this could be provoked
1184 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1185 components; fortuitously, this particular loophole had already been fixed
1186 by change 4.50/55 above.
1188 If there are any other similar loopholes, the new check in host_aton()
1189 itself should stop them being exploited. The report I received stated that
1190 data on the command line could provoke the exploit when Exim was running as
1191 exim, but did not say which command line option was involved. All I could
1192 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1193 running as the user.
ea3a6f44 1194 CAN-2005-0021
119661. There was a buffer overflow vulnerability in the SPA authentication code
1197 (which came originally from the Samba project). I have added a test to the
1198 spa_base64_to_bits() function which I hope fixes it.
ea3a6f44 1199 CAN-2005-0022
b975ba52 1200
120162. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1202 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
120463. The daemon start-up calls getloadavg() while still root for those OS that
1205 need the first call to be done as root, but it missed one case: when
1206 deliver_queue_load_max is set with deliver_drop_privilege. This is
1207 necessary for the benefit of the queue runner, because there is no re-exec
1208 when deliver_drop_privilege is set.
121064. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1211 This has been fixed.
121365. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1214 was in use, was not putting the data itself into the right store pool;
1215 consequently, it could be overwritten for a subsequent message in the same
1216 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1217 the caching.)
121966. Added hosts_max_try_hardlimit to the smtp transport, default 50.
122167. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1222 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1223 the function were treating the return as a boolean value, which happened to
1224 work because 0=false and not-0=true, but is not correct code.
122668. The host_aton() function was not handling scoped IPv6 addresses (those
1227 with, for example, "%eth0" on the end) correctly.
122969. Fixed some compiler warnings in acl.c for the bitmaps specified with
1230 negated items (that is, ~something) in unsigned ints. Some compilers
1231 apparently mutter when there is no cast.
123370. If an address verification called from an ACL failed, and did not produce a
1234 user-specific message (i.e. there was only a "system" message), nothing was
1235 put in $acl_verify_message. In this situation, it now puts the system
1236 message there.
123871. Change 4.23/11 added synchronization checking at the start of an SMTP
1239 session; change 4.31/43 added the unwanted input to the log line - except
1240 that it did not do this in the start of session case. It now does.
124272. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1243 This is wrong and can cause the other end to generate a synchronization
1244 error if it is another Exim or anything else that does the synchronization
1245 check. A QUIT command is no longer sent after a timeout.
124773. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1248 during host lookups.
125074. The defer_ok option of callout verification was not working if it was used
1251 when verifying addresses in header lines, that is, for this case:
1253 verify = header_sender/callout=defer_ok
125575. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1256 those file descriptors could be used for SMTP connections. If anything
1257 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1258 could be sent to the SMTP client, causing chaos. The daemon now opens
1259 stdin, stdout, and stderr to /dev/null when it puts itself into the
1260 background.
126276. Arrange for output from Perl's "warn" command to be written to Exim's main
1263 log by default. The user can override this with suitable Perl magic.
126577. The use of log_message on a "discard" ACL verb, which is supposed to add to
1266 the log message when discard triggers, was not working for the DATA ACL or
1267 for the non-SMTP ACL.
126978. Error message wording change in sieve.c.
127179. If smtp_accept_max_per_host was set, the number of connections could be
1272 restricted to fewer than expected, because the daemon was trying to set up
1273 a new connection before checking whether the processes handling previous
1274 connections had finished. The check for completed processes is now done
1275 earlier. On busy systems, this bug wouldn't be noticed because something
1276 else would have woken the daemon, and it would have reaped the completed
1277 process earlier.
127980. If a message was submitted locally by a user whose login name contained one
1280 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1281 It caused a spool format error. I have fixed the spool reading code. A
1282 related problem was that the "from" clause in the Received: line became
1283 illegal because of the space(s). It is now covered by ${quote_local_part.
128581. Included the latest eximstats from Steve (adds average sizes to HTML Top
1286 tables).
128882. Updated OS/Makefile-AIX as per message from Mike Meredith.
129083. Patch from Sieve maintainer to fix unterminated string problem in
1291 "vacation" handling.
129384. Some minor changes to the Linux configuration files to help with other
1294 OS variants using glibc.
129685. One more patch for Sieve to update vacation handling to latest spec.
495ae4b0 1298
1300See the note above about the 4.44 and 4.50 releases.
1304Exim version 4.44
1307 1. Change 4.43/35 introduced a bug that caused file counts to be
1308 incorrectly computed when quota_filecount was set in an appendfile
1309 transport
1311 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1312 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1313 place.
1315 3. Give more explanation in the error message when the command for a transport
1316 filter fails to execute.
1318 4. There are several places where Exim runs a non-Exim command in a
1319 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1320 was being done only for the command run by the queryprogram router. It is
1321 now done for all such subprocesses. The other cases are: ${run, transport
1322 filters, and the commands run by the lmtp and pipe transports.
1324 5. Some older OS have a limit of 256 on the maximum number of file
1325 descriptors. Exim was using setrlimit() to set 1000 as a large value
1326 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1327 systems. I've change it so that if it can't get 1000, it tries for 256.
1329 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1330 was an oversight, and furthermore, ever since the addition of extra
1331 controls (e.g. 4.43/32), the checks on when to allow different forms of
1332 "control" were broken. There should now be diagnostics for all cases when a
1333 control that does not make sense is encountered.
1335 7. $recipients is now available in the predata ACL (oversight).
1337 8. Tidy the search cache before the fork to do a delivery from a message
1338 received from the command line. Otherwise the child will trigger a lookup
1339 failure and thereby defer the delivery if it tries to use (for example) a
1340 cached ldap connection that the parent has called unbind on.
1342 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1343 of $address_data from the recipient verification was clobbered by the
1344 sender verification.
134610. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1347 was its contents. (It was OK if the option was not defined at all.)
134911. A "Completed" log line is now written for messages that are removed from
1350 the spool by the -Mrm option.
135212. $host_address is now set to the target address during the checking of
1353 ignore_target_hosts.
135513. When checking ignore_target_hosts for an ipliteral router, no host name was
1356 being passed; this would have caused $sender_host_name to have been used if
1357 matching the list had actually called for a host name (not very likely,
1358 since this list is usually IP addresses). A host name is now passed as
1359 "[x.x.x.x]".
136114. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1362 code that specifies a non-restarting handler (typically sigaction() in
1363 modern systems) in an attempt to fix a rare and obscure crash bug.
136515. Narrowed the window for a race in the daemon that could cause it to ignore
1366 SIGCHLD signals. This is not a major problem, because they are used only to
1367 wake it up if nothing else does.
136916. A malformed maildirsize file could cause Exim to calculate negative values
1370 for the mailbox size or file count. Odd effects could occur as a result.
1371 The maildirsize information is now recalculated if the size or filecount
1372 end up negative.
137417. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1375 support for a long time. Removed HAVE_SYS_VFS_H.
ea3a6f44 137718. Updated exipick to current release from John Jetmore.
137919. Allow an empty sender to be matched against a lookup in an address list.
1380 Previously the only cases considered were a regular expression, or an
1381 empty pattern.
138320. Exim went into a mad DNS lookup loop when doing a callout where the
1384 host was specified on the transport, if the DNS lookup yielded more than
1385 one IP address.
138721. The RFC2047 encoding function was originally intended for short strings
1388 such as real names; it was not keeping to the 75-character limit for
1389 encoded words that the RFC imposes. It now respects the limit, and
1390 generates multiple encoded words if necessary. To be on the safe side, I
1391 have increased the buffer size for the ${rfc2047: expansion operator from
1392 1024 to 2048 bytes.
bbe902f0 1393
139422. Failure to deliver a bounce message always caused it to be frozen, even if
1395 there was an errors_to setting on the router. The errors_to setting is now
1396 respected.
139823. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1399 canonical form (fully expanded) before being placed in
1400 $sender_host_address.
140224. Updated eximstats to version 1.33
140425. Include certificate and key file names in error message when GnuTLS fails
1405 to set them up, because the GnuTLS error message doesn't include the name
1406 of the failing file when there is a problem reading it.
140826. Expand error message when OpenSSL has problems setting up cert/key files.
ea3a6f44 1409 As per change 25.
bbe902f0 1410
141127. Reset the locale to "C" after calling embedded Perl, in case it was changed
1412 (this can affect the format of dates).
bbe902f0 1413
141428. exim_tidydb, when checking for the continued existence of a message for
1415 which it has found a message-specific retry record, was not finding
1416 messages that were in split spool directories. Consequently, it was
1417 deleting retry records that should have stayed in existence.
141929. eximstats updated to version 1.35
1420 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1421 1.35 - bugfix such that pie charts by volume are generated correctly
142330. The SPA authentication driver was not abandoning authentication and moving
1424 on to the next authenticator when an expansion was forced to fail,
1425 contradicting the general specification for all authenticators. Instead it
1426 was generating a temporary error. It now behaves as specified.
bbe902f0 1427
142831. The default ordering of permitted cipher suites for GnuTLS was pessimal
1429 (the order specifies the preference for clients). The order is now AES256,
1430 AES128, 3DES, ARCFOUR128.
bbe902f0 1431
143231. Small patch to Sieve code - explicitly set From: when generating an
1433 autoreply.
bbe902f0 1434
143532. Exim crashed if a remote delivery caused a very long error message to be
1436 recorded - for instance if somebody sent an entire SpamAssassin report back
1437 as a large number of 550 error lines. This bug was coincidentally fixed by
1438 increasing the size of one of Exim's internal buffers (big_buffer) that
1439 happened as part of the Exiscan merge. However, to be on the safe side, I
1440 have made the code more robust (and fixed the comments that describe what
1441 is going on).
bbe902f0 1442
144333. Some experimental protocols are using DNS PTR records for new purposes. The
1444 keys for these records are domain names, not reversed IP addresses. The
1445 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1446 leaves it alone. Component reversal etc. now happens only for IP addresses.
1447 CAN-2005-0021
144934. The host_aton() function is supposed to be passed a string that is known
1450 to be a valid IP address. However, in the case of IPv6 addresses, it was
1451 not checking this. This is a hostage to fortune. Exim now panics and dies
1452 if the condition is not met. A case was found where this could be provoked
1453 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1454 components; fortuitously, this particular loophole had already been fixed
1455 by change 4.50/55 or 4.44/33 above.
1457 If there are any other similar loopholes, the new check in host_aton()
1458 itself should stop them being exploited. The report I received stated that
1459 data on the command line could provoke the exploit when Exim was running as
1460 exim, but did not say which command line option was involved. All I could
1461 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1462 running as the user.
1463 CAN-2005-0021
146535. There was a buffer overflow vulnerability in the SPA authentication code
1466 (which came originally from the Samba project). I have added a test to the
1467 spa_base64_to_bits() function which I hope fixes it.
1468 CAN-2005-0022
147036. The daemon start-up calls getloadavg() while still root for those OS that
1471 need the first call to be done as root, but it missed one case: when
1472 deliver_queue_load_max is set with deliver_drop_privilege. This is
1473 necessary for the benefit of the queue runner, because there is no re-exec
1474 when deliver_drop_privilege is set.
bbe902f0 1475
147637. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1477 was in use, was not putting the data itself into the right store pool;
1478 consequently, it could be overwritten for a subsequent message in the same
1479 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1480 the caching.)
bbe902f0 1481
148238. Sometimes the final signoff response after QUIT could fail to get
1483 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1484 < 0 before doing a fflush(). This bug looks as though it goes back to the
1485 introduction of TLS in release 3.20, but "sometimes" must have been rare
1486 because the tests only now provoked it.
1489Exim version 4.43
1492 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1493 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1494 Now it returns an integer. A number of calls were still expecting a T/F
1495 return. Fortuitously, in all cases, the tests worked in OK situations,
1496 which is the norm. However, things would have gone wrong on any write
1497 failures on the smtp file descriptor. This function is used when sending
1498 messages over SMTP and also when doing verify callouts.
1500 2. When Exim is called to do synchronous delivery of a locally submitted
1501 message (the -odf or -odi options), it no longer closes stderr before doing
1502 the delivery.
1504 3. Implemented the mua_wrapper option.
1506 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1508 5. Implemented the functions header_remove(), header_testname(),
1509 header_add_at_position(), and receive_remove_recipient(), and exported them
1510 to local_scan().
1512 6. If an ACL "warn" statement specified the addition of headers, Exim already
1513 inserted X-ACL-Warn: at the start if there was no header name. However, it
1514 was not making this test for the second and subsequent header lines if
1515 there were newlines in the string. This meant that an invalid header could
1516 be inserted if Exim was badly configured.
1518 7. Allow an ACL "warn" statement to add header lines at the start or after all
1519 the Received: headers, as well as at the end.
1521 8. Added the rcpt_4xx retry error code.
1523 9. Added postmaster_mailfrom=xxx to callout verification option.
152510. Added mailfrom=xxxx to the callout verification option, for verify=
1526 header_sender only.
152811. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1529 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
153112. Inserted some casts to stop certain compilers warning when using pointer
1532 differences as field lengths or precisions in printf-type calls (mostly
1533 affecting debugging statements).
153513. Added optional readline() support for -be (dynamically loaded).
153714. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1538 same clock tick as a message's arrival, so that its received time was the
1539 same as the "first fail" time on the retry record, and that message
1540 remained on the queue past the ultimate address timeout, every queue runner
1541 would try a delivery (because it was past the ultimate address timeout) but
1542 after another failure, the ultimate address timeout, which should have then
1543 bounced the address, did not kick in. This was a "< instead of <=" error;
1544 in most cases the first failure would have been in the next clock tick
1545 after the received time, and all would be well.
154715. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1548 being recognized when the domain list was tested by the match_domain
1549 condition in an expansion string.
155116. Added the ${str2b64: operator.
155317. Exim was always calling setrlimit() to set a large limit for the number of
1554 processes, without checking whether the existing limit was already
1555 adequate. (It did check for the limit on file descriptors.) Furthermore,
1556 errors from getrlimit() and setrlimit() were being ignored. Now they are
1557 logged to the main and panic logs, but Exim does carry on, to try to do its
1558 job under whatever limits there are.
156018. Imported PCRE 5.0.
156219. Trivial typo in log message " temporarily refused connection" (the leading
1563 space).
156520. If the log selector return_path_on_delivery was set and an address was
1566 redirected to /dev/null, the delivery process crashed because it assumed
1567 that a return path would always be set for a "successful" delivery. In this
1568 case, the whole delivery is bypassed as an optimization, and therefore no
1569 return path is set.
157121. Internal re-arrangement: the function for sending a challenge and reading
1572 a response while authentication was assuming a zero-terminated challenge
1573 string. It's now changed to take a pointer and a length, to allow for
1574 binary data in such strings.
157622. Added the cyrus_sasl authenticator (code supplied by MBM).
157823. Exim was not respecting finduser_retries when seeking the login of the
1579 uid under which it was called; it was always trying 10 times. (The default
1580 setting of finduser_retries is zero.) Also, it was sleeping after the final
1581 failure, which is pointless.
158324. Implemented tls_on_connect_ports.
158525. Implemented acl_smtp_predata.
158726. If the domain in control=submission is set empty, Exim assumes that the
1588 authenticated id is a complete email address when it generates From: or
1589 Sender: header lines.
159127. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1592 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1593 chown and chgrp in /bin and hostname in /usr/bin.
159528. Exim was keeping the "process log" file open after each use, just as it
1596 does for the main log. This opens the possibility of it remaining open for
1597 long periods when the USR1 signal hits a daemon. Occasional processlog
1598 errors were reported, that could have been caused by this. Anyway, it seems
1599 much more sensible not to leave this file open at all, so that is what now
1600 happens.
160229. The long-running daemon process does not normally write to the log once it
1603 has entered its main loop, and it closes the log before doing so. This is
1604 so that log files can straightforwardly be renamed and moved. However,
1605 there are a couple of unusual error situations where the daemon does write
1606 log entries, and I had neglected to close the log afterwards.
160830. The text of an SMTP error response that was received during a remote
1609 delivery was being truncated at 512 bytes. This is too short for some of
1610 the long messages that one sometimes sees. I've increased the limit to
1611 1024.
161331. It is now possible to make retry rules that apply only when a message has a
1614 specific sender, in particular, an empty sender.
161632. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1617 it possible to be selective about when SMTP synchronization is enforced.
161933. Added "control = caseful_local_part" and "control = "caselower_local_part".
162132. Implemented hosts_connection_nolog.
162333. Added an ACL for QUIT.
162534. Setting "delay_warning=" to disable warnings was not working; it gave a
1626 syntax error.
162835. Added mailbox_size and mailbox_filecount to appendfile.
163036. Added control = no_multiline_responses to ACLs.
163237. There was a bug in the logic of the code that waits for the clock to tick
1633 in the case where the clock went backwards by a substantial amount such
1634 that the microsecond fraction of "now" was more than the microsecond
1635 fraction of "then" (but the whole seconds number was less).
163738. Added support for the libradius Radius client library this is found on
1638 FreeBSD (previously only the radiusclient library was supported).
1641Exim version 4.42
1644 1. When certain lookups returned multiple values in the form name=value, the
1645 quoting of the values was not always being done properly. Specifically:
1646 (a) If the value started with a double quote, but contained no whitespace,
1647 it was not quoted.
1648 (b) If the value contained whitespace other than a space character (i.e.
1649 tabs or newlines or carriage returns) it was not quoted.
1650 This fix has been applied to the mysql and pgsql lookups by writing a
1651 separate quoting function and calling it from the lookup code. The fix
1652 should probably also be applied to nisplus, ibase and oracle lookups, but
1653 since I cannot test any of those, I have not disturbed their existing code.
1655 2. A hit in the callout cache for a specific address caused a log line with no
1656 reason for rejecting RCPT. Now it says "Previous (cached) callout
1657 verification failure".
1659 3. There was an off-by-one bug in the queryprogram router. An over-long
1660 return line was truncated at 256 instead of 255 characters, thereby
1661 overflowing its buffer with the terminating zero. As well as fixing this, I
1662 have increased the buffer size to 1024 (and made a note to document this).
1664 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1665 when Exim is waiting for an SMTP response from a remote server, Exim
1666 restarts its select() call on the socket, thereby resetting its timeout.
1667 This is not a problem when such interrupts are rare. Somebody set up a cron
1668 job to run exiwhat every 2 minutes, which is less than the normal select()
1669 timeout (5 or 10 minutes). This meant that the select() timeout never
1670 kicked in because it was always reset. I have fixed this by comparing the
1671 time when an interrupt arrives with the time at the start of the first call
1672 to select(). If more time than the timeout has elapsed, the interrupt is
1673 treated as a timeout.
1675 5. Some internal re-factoring in preparation for the addition of Sieve
1676 extensions (by MH). In particular, the "personal" test is moved to a
1677 separate function, and given an option for scanning Cc: and Bcc: (which is
1678 not set for Exim filters).
1680 6. When Exim created an email address using the login of the caller as the
1681 local part (e.g. when creating a From: or Sender: header line), it was not
1682 quoting the local part when it contained special characters such as @.
1684 7. Installed new OpenBSD configuration files.
1686 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1687 try to make them clearer.
1689 9. Callout options, other than the timeout value, were being ignored when
1690 verifying sender addresses in header lines. For example, when using
1692 verify = header_sender/callout=no_cache
1694 the cache was (incorrectly) being used.
169610. Added a missing instance of ${EXE} to the exim_install script; this affects
1697 only the Cygwin environment.
169911. When return_path_on_delivery was set as a log selector, if different remote
1700 addresses in the same message used different return paths and parallel
1701 remote delivery occurred, the wrong values would sometimes be logged.
1702 (Whenever a remote delivery process finished, the return path value from
1703 the most recently started remote delivery process was logged.)
170512. RFC 3848 specifies standard names for the "with" phrase in Received: header
1706 lines when AUTH and/or TLS are in use. This is the "received protocol"
1707 field. Exim used to use "asmtp" for authenticated SMTP, without any
1708 indication (in the protocol name) for TLS use. Now it follows the RFC and
1709 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1710 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1711 names appear in log lines as well as in Received: header lines.
171313. Installed MH's patches for Sieve to add the "copy" and "vacation"
1714 extensions, and comparison tests, and to fix some bugs.
171614. Changes to the "personal" filter test:
1718 (1) The test was buggy in that it was just doing the equivalent of
1719 "contains" tests on header lines. For example, if a user's address was
1720 anne@some.where, the "personal" test would incorrectly be true for
1722 To: susanne@some.where
1724 This test is now done by extracting each address from the header in turn,
1725 and checking the entire address. Other tests that are part of "personal"
1726 are now done using regular expressions (for example, to check local parts
1727 of addresses in From: header lines).
1729 (2) The list of non-personal local parts in From: addresses has been
1730 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1731 taken from the Sieve specification recommendations.
1733 (3) If the message contains any header line starting with "List-" it is
1734 treated as non-personal.
1736 (4) The test for "circular" in the Subject: header line has been removed
1737 because it now seems ill-conceived.
173915. Minor typos in src/EDITME comments corrected.
174116. Installed latest exipick from John Jetmore.
174317. If headers_add on a router specified a text string that was too long for
1744 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1745 of string_sprintf() is now avoided.
174718. $message_body_size was not set (it was always zero) when running the DATA
1748 ACL and the local_scan() function.
175019. For the "mail" command in an Exim filter, no default was being set for
1751 the once_repeat time, causing a random time value to be used if "once" was
1752 specified. (If the value happened to be <= 0, no repeat happened.) The
1753 default is now 0s, meaning "never repeat". The "vacation" command was OK
1754 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1755 (I found it when inspecting the code).
175720. There is now an overall timeout for performing a callout verification. It
1758 defaults to 4 times the callout timeout, which applies to individual SMTP
1759 commands during the callout. The overall timeout applies when there is more
1760 than one host that can be tried. The timeout is checked before trying the
1761 next host. This prevents very long delays if there are a large number of
1762 hosts and all are timing out (e.g. when the network connections are timing
1763 out). The value of the overall timeout can be changed by specifying an
1764 additional sub-option for "callout", called "maxwait". For example:
1766 verify = sender/callout=5s,maxwait=20s
176821. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1769 to the end before writing, but this should make it even safer).
177122. Exim was forgetting that it had advertised PIPELINING for the second and
1772 subsequent messages on an SMTP connection. It was also not resetting its
1773 memory on STARTTLS and an internal HELO.
177523. When Exim logs an SMTP synchronization error within a session, it now
1776 records whether PIPELINING has been advertised or not.
177824. Added 3 instances of "(long int)" casts to time_t variables that were being
1779 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1780 rather than long int.
178225. Installed the latest Cygwin configuration files from the Cygwin maintainer.
178426. Added the never_mail option to autoreply.
1787Exim version 4.41
1790 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1791 crash if the getsockname() call failed; this can happen if a connection is
1792 closed very soon after it is established. The problem was simply in the
1793 order in which certain operations were done, causing Exim to try to write
1794 to the SMTP stream before it had set up the file descriptor. The bug has
1795 been fixed by making things happen in the correct order.
1798Exim version 4.40
1801 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1802 before the connection was closed, thus losing the rejection response.
1804 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1805 some early Solaris releases, but causes trouble in current releases where
1806 socklen_t is defined.
1808 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1809 always exist.
1811 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1812 configured.
1814 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1815 that releasing the top of it at the end releases what was used for sub-
1816 expansions (unless the block got too big). However, discard this block if
1817 the first thing is a variable or header, so that we can use its block when
1818 it is dynamic (useful for very large $message_headers, for example).
1820 6. Lookups now cache *every* query, not just the most recent. A new, separate
1821 store pool is used for this. It can be recovered when all lookup caches are
1822 flushed. Lookups now release memory at the end of their result strings.
1823 This has involved some general refactoring of the lookup sources.
1825 7. Some code has been added to the store_xxx() functions to reduce the amount
1826 of flapping under certain conditions.
1828 8. log_incoming_interface used to affect only the <= reception log lines. Now
1829 it causes the local interface and port to be added to several more SMTP log
1830 lines, for example "SMTP connection from", and rejection lines.
1832 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
183410. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
183611. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1837 could be overwritten at the end of the current message (or the start of a
1838 new message if it was set in a HELO ACL). The value is now preserved for
1839 the duration of the SMTP connection.
184112. If a transport had a headers_rewrite setting, and a matching header line
1842 contained an unqualified address, that address was qualified, even if it
1843 did not match any rewriting rules. The underlying bug was that the values
1844 of the flags that permit the existence of unqualified sender and recipient
1845 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1846 non-local messages, and by -bnq for local messages) were not being
1847 preserved with the message after it was received.
184913. When Exim was logging an SMTP synchronization error, it could sometimes log
1850 "next input=" as part of the text comprising the host identity instead of
1851 the correct text. The code was using the same buffer for two different
1852 strings. However, depending on which order the printing function evaluated
1853 its arguments, the bug did not always show up. Under Linux, for example, my
1854 test suite worked just fine.
185614. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1857 doesn't work with some older versions of Perl. It has been changed to "my",
1858 which in any case is probably the better facility to use.
186015. A really picky compiler found some instances of statements for creating
1861 error messages that either had too many or two few arguments for the format
1862 string.
186416. The size of the buffer for calls to the DNS resolver has been increased
1865 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1866 for addresses that have a lot of PTR records. This alleviates a problem; it
1867 does not fully solve it.
186917. A dnsdb lookup for PTR records that receives more data than will fit in the
1870 buffer now truncates the list and logs the incident, which is the same
1871 action as happens when Exim is looking up a host name and its aliases.
1872 Previously in this situation something unpredictable would happen;
1873 sometimes it was "internal error: store_reset failed".
187518. If a server dropped the connection unexpectedly when an Exim client was
1876 using GnuTLS and trying to read a response, the client delivery process
1877 crashed while trying to generate an error log message.
187919. If a "warn" verb in an ACL added multiple headers to a message in a single
1880 string, for example:
1882 warn message = H1: something\nH2: something
1884 the text was added as a single header line from Exim's point of view
1885 though it ended up OK in the delivered message. However, searching for the
1886 second and subsequent header lines using $h_h2: did not work. This has been
1887 fixed. Similarly, if a system filter added multiple headers in this way,
1888 the routers could not see them.
189020. Expanded the error message when iplsearch is called with an invalid key to
1891 suggest using net-iplsearch in a host list.
189321. When running tests using -bh, any delays imposed by "delay" modifiers in
1894 ACLs are no longer actually imposed (and a message to that effect is
1895 output).
189722. If a "gecos" field in a passwd entry contained escaped characters, in
1898 particular, if it contained a \" sequence, Exim got it wrong when building
1899 a From: or a Sender: header from that name. A second bug also caused
1900 incorrect handling when an unquoted " was present following a character
1901 that needed quoting.
190323. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1904 was not being matched caselessly.
190624. Arranged for all hyphens in the exim.8 source to be escaped with
1907 backslashes.
190925. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1910 properly. Recipient callout cache records were still being keyed to include
1911 the sender, even when use_sender was set false. This led to far more
1912 callouts that were necessary. The sender is no longer included in the key
1913 when use_sender is false.
191526. Added "control = submission" modifier to ACLs.
191727. Added the ${base62d: operator to decode base 62 numbers.
191928. dnsdb lookups can now access SRV records.
192129. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1922 the configuration file.
192430. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1925 (-v) mode. This makes the output for a verbose queue run more intelligible.
192731. Added a use_postmaster feature to recipient callouts.
192932. Added the $body_zerocount variable, containing the number of binary zero
1930 bytes in the message body.
193233. The time of last modification of the "new" subdirectory is now used as the
1933 "mailbox time last read" when there is a quota error for a maildir
1934 delivery.
193634. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
193835. Added +ignore_unknown as a special item in host lists.
194036. Code for decoding IPv6 addresses in host lists is now included, even if
1941 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1942 address was recognized as an IP address, but was then not correctly decoded
1943 into binary, causing unexpected and incorrect effects when compared with
1944 another IP address.
1947Exim version 4.34
1950 1. Very minor rewording of debugging text in manualroute to say "list of
1951 hosts" instead of "hostlist".
1953 2. If verify=header_syntax was set, and a header line with an unqualified
1954 address (no domain) and a large number of spaces between the end of the
1955 name and the colon was received, the reception process suffered a buffer
1956 overflow, and (when I tested it) crashed. This was caused by some obsolete
1957 code that should have been removed. The fix is to remove it!
1959 3. When running in the test harness, delay a bit after writing a bounce
1960 message to get a bit more predictability in the log output.
1962 4. Added a call to search_tidyup() just before forking a reception process. In
1963 theory, someone could use a lookup in the expansion of smtp_accept_max_
1964 per_host which, without the tidyup, could leave open a database connection.
1966 5. Added the variables $recipient_data and $sender_data which get set from a
1967 lookup success in an ACL "recipients" or "senders" condition, or a router
1968 "senders" option, similar to $domain_data and $local_part_data.
1970 6. Moved the writing of debug_print from before to after the "senders" test
1971 for routers.
1973 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1974 problems for message scanning, either using a data ACL, or using
1975 local_scan() because the Received: header was not generated till after they
1976 were called (in order to set the time as the time of reception completion).
1977 I have revised the way this works. The header is now generated after the
1978 body is received, but before the ACL or local_scan() are called. After they
1979 are run, the timestamp in the header is updated.
1982Exim version 4.33
1985 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1986 before starting a queue runner without re-exec. This happened only when
1987 deliver_drop_privilege was set or when the Exim user was set to root. The
1988 effect of the bug was that timeouts during subsequent deliveries caused
1989 crashes instead of being properly handled. The handler is now left at its
1990 default (and expected) setting.
1992 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1993 message, again when deliver_drop_privilege is set or Exim is run as root.
1994 The bug described in (1) was not present in this case, but the tidying up
1995 of the other signals was missing. I have made the two cases consistent.
1997 3. The ignore_target_hosts setting on a manualroute router was being ignored
1998 for hosts that were looked up using the /MX notation.
2000 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2001 in domain lists.
2003 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2004 operated on the sender address. After changing the $sender_address to <>
2005 for the sender address verify, Exim was re-instated it as the original
2006 (before rewriting) address, but remembering that it had rewritten it, so it
2007 wasn't rewriting it again. This bug also had the effect of breaking the
2008 sender address verification caching when the sender address was rewritten.
2010 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2011 This has been changed so that if the ip literal address matches
2012 ignore_target_hosts, the router declines.
2014 7. Added expansion conditions match_domain, match_address, and match_local_
2015 part (NOT match_host).
2017 8. The placeholder for the Received: header didn't have a length field set.
2019 9. Added code to Exim itself and to exim_lock to test for a specific race
2020 condition that could lead to file corruption when using MBX delivery. The
2021 issue is with the lockfile that is created in /tmp. If this file is removed
2022 after a process has opened it but before that process has acquired a lock,
2023 there is the potential for a second process to recreate the file and also
2024 acquire a lock. This could lead to two Exim processes writing to the file
2025 at the same time. The added code performs the same test as UW imapd; it
2026 checks after acquiring the lock that its file descriptor still refers to
2027 the same named file.
202910. The buffer for building added header lines was of fixed size, 8192 bytes.
2030 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2031 when Exim is built.
203311. Added the smtp_active_hostname option. If used, this will typically be made
2034 to depend on the incoming interface address. Because $interface_address is
2035 not set up until the daemon has forked a reception process, error responses
2036 that can happen earlier (such as "too many connections") no longer contain
2037 a host name.
203912. If an expansion in a condition on a "warn" statement fails because a lookup
2040 defers, the "warn" statement is abandoned, and the next ACL statement is
2041 processed. Previously this caused the whole ACL to be aborted.
204313. Added the iplsearch lookup type.
204514. Added ident_timeout as a log selector.
204715. Added tls_certificate_verified as a log selector.
204916. Added a global option tls_require_ciphers (compare the smtp transport
2050 option of the same name). This controls incoming TLS connections.
205217. I finally figured out how to make tls_require_ciphers do a similar thing
2053 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2054 before starting the TLS session.
205618. Tabs are now shown as \t in -bP output.
205819. If the log selector return_path_on_delivery was set, Exim crashed when
2059 bouncing a message because it had too many Received: header lines.
206120. If two routers both had headers_remove settings, and the first one included
2062 a superfluous trailing colon, the final name in the first list and the
2063 first name in the second list were incorrectly joined into one item (with a
2064 colon in the middle).
2067Exim version 4.32
2070 1. Added -C and -D options to the exinext utility, mainly to make it easier
2071 to include in the automated testing, but these could be helpful when
2072 multiple configurations are in use.
2074 2. The exinext utility was not formatting the output nicely when there was
2075 an alternate port involved in the retry record key, nor when there was a
2076 message id as well (for retries that were specific to a specific message
2077 and a specific host). It was also confused by IPv6 addresses, because of
2078 the additional colons they contain. I have fixed the IPv4 problem, and
2079 patched it up to do a reasonable job for IPv6.
2081 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2082 delivery, the log line now contains "pipelined" if PIPELINING was used.
2084 4. An SMTP transport process used to panic and die if the bind() call to set
2085 an explicit outgoing interface failed. This has been changed; it is now
2086 treated in the same way as a connect() failure.
2088 5. A reference to $sender_host_name in the part of a conditional expansion
2089 that was being skipped was still causing a DNS lookup. This no longer
2090 occurs.
2092 6. The def: expansion condition was not recognizing references to header lines
2093 that used bh_ and bheader_.
2095 7. Added the _cache feature to named lists.
2097 8. The code for checking quota_filecount in the appendfile transport was
2098 allowing one more file than it should have been.
2100 9. For compatibility with Sendmail, the command line option
2102 -prval:sval
2104 is equivalent to
2106 -oMr rval -oMs sval
2108 and sets the incoming protocol and host name (for trusted callers). The
2109 host name and its colon can be omitted when only the protocol is to be set.
2110 Note the Exim already has two private options, -pd and -ps, that refer to
2111 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2112 "s", but I don't think that's a major issue.
211410. A number of refactoring changes to the code, none of which should affect
2115 Exim's behaviour:
2117 (a) The number of logging options was getting close to filling up the
2118 32-bit word that was used as a bit map. I have split them into two classes:
2119 those that are passed in the argument to log_write(), and those that are
2120 only ever tested independently outside of that function. These are now in
2121 separate 32-bit words, so there is plenty of room for expansion again.
2122 There is no change in the user interface or the logging behaviour.
2124 (b) When building, for example, log lines, the code previously used a
2125 macro that called string_cat() twice, in order to add two strings. This is
2126 not really sufficiently general. Furthermore, there was one instance where
2127 it was actually wrong because one of the argument was used twice, and in
2128 one call a function was used. (As it happened, calling the function twice
2129 did not affect the overall behaviour.) The macro has been replaced by a
2130 function that can join an arbitrary number of extra strings onto a growing
2131 string.
2133 (c) The code for expansion conditions now uses a table and a binary chop
2134 instead of a serial search (which was left over from when there were very
2135 few conditions). Also, it now recognizes conditions like "pam" even when
2136 the relevant support is not compiled in: a suitably worded error message is
2137 given if an attempt is made to use such a condition.
213911. Added ${time_interval:xxxxx}.
214112. A bug was causing one of the ddress fields not to be passed back correctly
2142 from remote delivery subprocesses. The field in question was not being
2143 subsequently used, so this caused to problems in practice.
214513. Added new log selectors queue_time and deliver_time.
214714. Might have fixed a bug in maildirsizefile handling that threw up
2148 "unexpected character" debug warnings, and recalculated the data
2149 unnecessarily. In any case, I expanded the warning message to give more
2150 information.
215215. Added the message "Restricted characters in address" to the statements in
2153 the default ACL that block characters like @ and % in local parts.
215516. Change 71 for release 4.31 proved to be much less benign that I imagined.
2156 Three changes have been made:
2158 (a) There was a serious bug; a negative response to MAIL caused the whole
2159 recipient domain to be cached as invalid, thereby blocking all messages
2160 to all local parts at the same domain, from all senders. This bug has
2161 been fixed. The domain is no longer cached after a negative response to
2162 MAIL if the sender used is not empty.
2164 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2165 been restored.
2167 (c) A new callout option, "use_sender" has been added for people who want
2168 the modified behaviour.
2171Exim version 4.31
2174 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2175 Larry Rosenman.
2177 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2178 indeed breaks things for older releases.
2180 3. Added additional logging to the case where there is a problem reading data
2181 from a filter that is running in a subprocess using a pipe, in order to
2182 try to track down a specific problem.
2184 4. Testing facility fudge: when running in the test harness and attempting
2185 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2186 getting "No route to host". Convert this to a timeout.
2188 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2189 warning.
2191 6. Some OS don't have socklen_t but use size_t instead. This affects the
2192 fifth argument of getsockopt() amongst other things. This is now
2193 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2194 can be set for individual OS. I have set it for SunOS5, OSF1, and
2195 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2196 some earlier ones do not.
2198 7. Change 4.30/15 was not doing the test caselessly.
2200 8. The standard form for an IPv6 address literal was being rejected by address
2201 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2202 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2203 this, as well as the form without the "IPv6" on the front (but only when
2204 address literals are enabled, of course).
2206 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
220810. Exim crashed if a message with an empty sender address specified by -f
2209 encountered a router with an errors_to setting. This could be provoked only
2210 by a command such as
2212 exim -f "" ...
2214 where an empty string was supplied; "<>" did not hit this bug.
221611. Installed PCRE release 4.5.
221812. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2219 remained set. It is now erased.
222113. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2222 times from message ids (which are base 36 rather than the normal 62).
222414. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2225 were being counted as actual protocol errors, and logged if the log
2226 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2227 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2228 and DATA following a set of rejected RCPTs do not count as protocol errors.
2229 In other words, Exim assumes they were pipelined, though this may not
2230 actually be the case. Of course, in all cases the client gets an
2231 appropriate error code.
223315. If a lookup fails in an ACL condition, a message about the failure may
2234 be available; it is used if testing the ACL cannot continue, because most
2235 such messages specify what the cause of the deferral is. However, some
2236 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2237 that caused an old message to be retained and used if a later statement
2238 caused a defer, replacing the real cause of the deferral.
224016. If an IP address had so many PTR records that the DNS lookup buffer
2241 was not large enough to hold them, Exim could crash while trying to process
2242 the truncated data. It now detects and logs this case.
224417. Further to 4.21/58, another change has been made: if (and only if) the
2245 first line of a message (the first header line) ends with CRLF, a bare LF
2246 in a subsequent header line has a space inserted after it, so as not to
2247 terminate the header.
224918. Refactoring: tidied an ugly bit of code in appendfile that copied data
2250 unnecessarily, used atoi() instead of strtol(), and didn't check the
2251 termination when getting file sizes from file names by regex.
225319. Completely re-implemented the support for maildirsize files, in the light
2254 of a number of problems with the previous contributed implementation
2255 (4.30/29). In particular:
2257 . If the quota is zero, the maildirsize file is maintained, but no quota is
2258 imposed.
2260 . If the maildir directory does not exist, it is created before any attempt
2261 to write a maildirsize file.
2263 . The quota value in the file is just a cache; if the quota is changed in
2264 the transport, the new value overrides.
2266 . A regular expression is available for excluding directories from the
2267 count.
226920. The autoreply transport checks the characters in options that define the
2270 message's headers; it allows continued headers, but it was checking with
2271 isspace() after an embedded newline instead of explicitly looking for a
2272 space or a tab.
227421. If all the "regular" hosts to which an address was routed had passed their
2275 expiry times, and had not reached their retry times, the address was
2276 bounced, even if fallback hosts were defined. Now Exim should go on to try
2277 the fallback hosts.
227922. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2280 equivalent code in the SMTP transport. Some hosts send humungous responses
2281 to HELO/EHLO, more than 1024 it seems.
228323. Refactoring: code in filter.c used (void *) for "any old type" but this
2284 gives compiler warnings in some environments. I've now done it "properly",
2285 using a union.
228724. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2288 (because of problems with the built-in one) was declared to return uschar *
2289 instead of char *, causing compiler failure.
229125. Fixed a file descriptor leak when processing alias/forward files.
229326. Fixed a minor format string issue in dbfn.c.
229527. Typo in exim.c: ("dmbnz" for "dbmnz").
229728. If a filter file refered to $h_xxx or $message_headers, and the headers
2298 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2299 become corrupted.
230129. When a sender address is verified, it is cached, to save repeating the test
2302 when there is more than one recipient in a message. However, when the
2303 verification involves a callout, it is possible for different callout
2304 options to be set for different recipients. It is too complicated to keep
2305 track of this in the cache, so now Exim always runs a verification when a
2306 callout is required, relying on the callout cache for the optimization.
2307 The overhead is duplication of the address routing, but this should not be
2308 too great.
231030. Fixed a bug in callout caching. If a RCPT command caused the sender address
2311 to be verified with callout=postmaster, and the main callout worked but the
2312 postmaster check failed, the verification correctly failed. However, if a
2313 subsequent RCPT command asked for sender verification *without* the
2314 postmaster check, incorrect caching caused this verification also to fail,
2315 incorrectly.
231731. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2318 it was not caching the DNS options (qualify_single, search_parents) that
2319 were used when the lookup failed. A subsequent lookup with different
2320 options therefore always gave the same answer, though there were cases
2321 where it should not have. (Example: a "domains = !$mx_any" option on a
2322 dnslookup router: the "domains" option is always processed without any
2323 widening, but the router might have qualify_single set.) Now Exim uses the
2324 cached value only when the same options are set.
232632. Added John Jetmore's "exipick" utility to the distribution.
232833. GnuTLS: When an attempt to start a TLS session fails for any reason other
2329 than a timeout (e.g. a certificate is required, and is not provided), an
2330 Exim server now closes the connection immediately. Previously it waited for
2331 the client to close - but if the client is SSL, it seems that they each
2332 wait for each other, leading to a delay before one of them times out.
233434: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2335 maintained 0.8.x compatibility because I don't think many are using it, and
2336 it is clearly obsolete.
233835. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2339 transport.
234136. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2342 client certificate was expired. A simple patch fixes this, though I don't
2343 understand the full logic of why the verify callback is called multiple
2344 times.
234637. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2347 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2348 which causes problems with some clients (such as the Certicom SSL Plus
2349 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2350 disables the coutermeasure allowing Eudora to connect."
235238. Exim was not checking that a write() to a log file succeeded. This could
2353 lead to Bad Things if a log got too big, in particular if it hit a file
2354 size limit. Exim now panics and dies if it cannot write to a log file, just
2355 as it does if it cannot open a log file.
235739. Modified OS/Makefile-Linux so that it now contains
2361 The two -D definitions ensure that Exim is compiled with large file
2362 support, which makes it possible to handle log files that are bigger than
2363 2^31.
236540. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2366 instance) a domain was checked against a named list that involved a lookup,
2367 causing $domain_data to be set, then another domain was checked against the
2368 same list, then the first domain was re-checked, the value of $domain_data
2369 after the final check could be wrong. In particular, if the second check
2370 failed, it could be set empty. This bug probably also applied to
2371 $localpart_data.
237341. The strip_trailing_dot option was not being applied to the address given
2374 with the -f command-line option.
237642. The code for reading a message's header from the spool was incrementing
2377 $received_count, but never initializing it. This meant that the value was
2378 incorrect (doubled) while delivering a message in the same process in which
2379 it was received. In the most common configuration of Exim, this never
2380 happens - a fresh exec is done - but it can happen when
2381 deliver_drop_privilege is set.
238343. When Exim logs an SMTP synchronization error - client data sent too soon -
2384 it now includes up to 150 characters of the unexpected data in the log
2385 line.
238744. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2388 and building data strings. The size of both of these buffers was 10 000
2389 bytes - far larger than anybody would *ever* want, thought I. Needless to
2390 say, somebody hit the limit. I have increased the maximum line length to
2391 20 000 and the maximum data length of concatenated lines to 100 000. I have
2392 also fixed two bugs, because there was no checking on these buffers. Tsk,
2393 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2394 buffer is too small.
239645. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2397 lsearch lookups. Now it does.
239946. When parsing a route_list item in a manualroute router, a fixed-length
2400 buffer was used for the list of hosts. I made this 1024 bytes long,
2401 thinking that nobody would ever have a list of hosts that long. Wrong.
2402 Somebody had a whole pile of complicated expansion conditions, and the
2403 string was silently truncated, leading to an expansion error. It turns out
2404 that it is easier to change to an unlimited length (owing to other changes
2405 that have happened since this code was originally written) than to build
2406 structure for giving a limitation error. The length of the item that
2407 expands into the list of hosts is now unlimited.
240947. The lsearch lookup could not handle data where the length of text line was
2410 more than 4095 characters. Such lines were truncated, leading to shortened
2411 data being returned. It should now handle lines of any length.
241348. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2414 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2415 ACL").
241749. Cosmetic tidy to scripts like exicyclog that are generated by globally
2418 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2419 no longer happens in comment lines. A list of replacements is now placed
2420 at the head of all of the source files, except those whose only change is
2421 to replace PERL_COMMAND in the very first #! line.
242350. Replaced the slow insertion sort in queue.c, for sorting the list of
2424 messages on the queue, with a bottom-up merge sort, using code contributed
2425 by Michael Haardt. This should make operations like -bp somewhat faster on
2426 large queues. It won't affect queue runners, except when queue_run_in_order
2427 is set.
242951. Installed eximstats 1.31 in the distribution.
243152. Added support for SRV lookups to the dnslookup router.
243353. If an ACL referred to $message_body or $message_body_end, the value was not
2434 reset for any messages that followed in the same SMTP session.
243654. The store-handling optimization for building very long strings was not
2437 differentiating between the different store pools. I don't think this
2438 actually made any difference in practice, but I've tidied it.
244055. While running the routers to verify a sender address, $sender_address
2441 was still set to the sender address. This is wrong, because when routing to
2442 send a bounce to the sender, it would be empty. Therefore, I have changed
2443 it so that, while verifying a sender address, $sender_address is set to <>.
2444 (There is no change to what happens when verifying a recipient address.)
244656. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2447 target A or AAAA records (if not already returned) without resetting the
2448 qualify_single or search_parents options of the DNS resolver. These are
2449 inappropriate in this case because the targets of MX and SRV records must
2450 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2451 target that, when qualified, matched something in the local domain. These
2452 two options are now turned off when doing these lookups.
245457. It seems that at least some releases of Reiserfs (which does not have the
2455 concept of a fixed number of inodes) returns zero and not -1 for the
2456 number of available inodes. This interacted badly with check_spool_inodes,
2457 which assumed that -1 was the "no such thing" setting. What I have done is
2458 to check that the total number of inodes is greater than zero before doing
2459 the test of how many are available.
246158. When a "warn" ACL statement has a log_message modifier, the message is
2462 remembered, and not repeated. This is to avoid a lot of repetition when a
2463 message has many recipients that cause the same warning to be written.
2464 Howewer, Exim was preserving the list of already written lines for an
2465 entire SMTP session, which doesn't seem right. The memory is now reset if a
2466 new message is started.
246859. The "rewrite" debugging flag was not showing the result of rewriting in the
2469 debugging output unless log_rewrite was also set.
247160. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2472 of (int)(handle) when we know that handle contains (void *)(-1).
247461. The Exim daemon panic-logs an error return when it closes the incoming
2475 connection. However "connection reset by peer" seems to be common, and
2476 isn't really an error worthy of noting specially, so that particular error
2477 is no long logged.
247962. When Exim is trying to find all the local interfaces, it used to panic and
2480 die if the ioctl to get the interface flags failed. However, it seems that
2481 on at least one OS (Solaris 9) it is possible to have an interface that is
2482 included in the list of interfaces, but for which you get a failure error
2483 for this call. This happens when the interface is not "plumbed" into a
2484 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2485 failure of the "get flags" call assumes that the interface is down.
248763. Added a ${eval10: operator, which assumes all numbers are decimal. This
2488 makes life easier for people who are doing arithmetic on fields extracted
2489 from dates, where you often get leading zeros that should not be
2490 interpreted as octal.
249264. Added qualify_domain to the redirect router, to override the global
2493 setting.
249565. If a pathologically long header line contained very many addresses (the
2496 report of this problem mentioned 10 000) and each of them was rewritten,
2497 Exim could use up a very large amount of memory. (It kept on making new
2498 copies of the header line as it rewrote, and never released the old ones.)
2499 At the expense of a bit more processing, the header rewriting function has
2500 been changed so that it no longer eats memory in this way.
250266. The generation of the Received: header has been moved from the time that a
2503 message starts to be received, to the time that it finishes. The timestamp
2504 in the Received: header should now be very close to that of the <= log
2505 line. There are two side-effects of this change:
2507 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2508 logged header lines no longer include the local Received: line, because
2509 it has not yet been created. The same applies to a copy of the message
2510 that is returned to a non-SMTP sender when a message is rejected.
2512 (b) When a filter file is tested using -bf, no additional Received: header
2513 is added to the test message. After some thought, I decided that this
2514 is a bug fix.
2516 This change does not affect the value of $received_for. It is still set
2517 after address rewriting, but before local_scan() is called.
251967. Installed the latest Cygwin-specific files from the Cygwin maintainer.
252168. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2522 gave an unhelpful panic error message, and a defer error. I have managed to
2523 change this behaviour so that it now rejects any supplied certificate,
2524 which seems right, as the list of acceptable certificates is empty.
252669. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2527 gave an unhelpful defer error. I have not managed to make this reject any
2528 supplied certificates, but the error message it gives is "no certificate
2529 supplied", which is not helpful.
253170. exigrep's output now also includes lines that are not associated with any
2532 message, but which match the given pattern. Implemented by a patch from
2533 Martin Sluka, which also tidied up the Perl a bit.
253571. Recipient callout verification, like sender verification, was using <> in
2536 the MAIL FROM command. This isn't really the right thing, since the actual
2537 sender may affect whether the remote host accepts the recipient or not. I
2538 have changed it to use the actual sender in the callout; this means that
2539 the cache record is now keyed on a recipient/sender pair, not just the
2540 recipient address. There doesn't seem to be a real danger of callout loops,
2541 since a callout by the remote host to check the sender would use <>.
2542 [SEE ABOVE: changed after hitting problems.]
254472. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2545 temporary errors. However, in the case of such a code being given after
2546 the end of a data transmission (i.e. after ".") Exim was failing to write
2547 a retry record for the message. (Yes, there was some broken host that was
2548 actually sending 8xx at this point.)
255073. An unknown lookup type in a host list could cause Exim to panic-die when
2551 the list was checked. (An example that provoked this was putting <; in the
2552 middle of a list instead of at the start.) If this happened during a DATA
2553 ACL check, a -D file could be left lying around. This kind of configuration
2554 error no longer causes Exim to die; instead it causes a defer errror. The
2555 incident is still logged to the main and panic logs.
255774. Buglet left over from Exim 3 conversion. The message "too many messages
2558 in one connection" was written to the rejectlog but not the mainlog, except
2559 when address rewriting (yes!) was being logged.
256175. Added write_rejectlog option.
256376. When a system filter was run not as root (that is, when system_filter_user
2564 was set), the values of the $n variables were not being returned to the
2565 main process; thus, they were not subsequently available in the $sn
2566 variables.
256877. Added +return_path_on_delivery log selector.
257078. A connection timeout was being treated differently from recipients deferred
2571 when testing hosts_max_try with a message that was older than the host's
2572 retry timeout. (The host should not be counted, thus allowing all hosts to
2573 be tried at least once before bouncing.) This may have been the cause of an
2574 occasionally reported bug whereby a message would remain on the queue
2575 longer than the retry timeout, but would be bounced if a delivery was
2576 forced. I say "may" because I never totally pinned down the problem;
2577 setting up timeout/retry tests is difficult. See also the next item.
257979. The ultimate address timeout was not being applied to errors that involved
2580 a combination of host plus message (for example, a timeout on a MAIL
2581 command). When an address resolved to a number of possible hosts, and they
2582 were not all tried for each delivery (e.g. because of hosts_max_try), a
2583 message could remain on the queue longer than the retry timeout.
258580. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2586 Haardt.
258881. Fixed an obscure SMTP outgoing bug which required at least the following
2589 conditions: (a) there was another message waiting for the same server;
2590 (b) the server returned 5xx to all RCPT commands in the first message so
2591 that the message was not completed; (c) the server dropped the connection
2592 or gave a negative response to the RSET that Exim sends to abort the
2593 transaction. The observed case was a dropped connection after DATA that had
2594 been sent in pipelining mode. That is, the server had advertised PIPELINING
2595 but was not implementing it correctly. The effect of the bug was incorrect
2596 behaviour, such as trying another host, and this could lead to a crash.
2599Exim version 4.30
2602 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2603 and daemon.c were passed as pointers to ints; they should have been
2604 pointers to socklen_t variables (which are typically unsigned ints).
2606 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2607 fixed.
2609 3. Fixed a really odd bug that affected only the testing scheme; patching a
2610 certain fixed string in the binary changed the value of another string that
2611 happened to be identical to the end of the original first string.
2613 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2614 name", it returns that address as the IP address. On some operating
2615 systems (e.g. Solaris), it also passes back the IP address string as the
2616 "host name". However, on others (e.g. Linux), it passes back an empty
2617 string. Exim wasn't checking for this, and was changing the host name to an
2618 empty string, assuming it had been canonicized.
2620 5. Although rare, it is permitted to have more than one PTR record for a given
2621 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2622 all the names associated with an address, because they do in Solaris.
2623 However, it seems that they do not in Linux for data that comes from the
2624 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2625 I found this out when I moved to a new Linux workstation and tried to run
2626 the Exim test suite.
2628 To get round this problem I have changed the code so that it now does its
2629 own call to the DNS to look up PTR records when searching for a host name.
2630 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2631 addresses that are only in /etc/hosts are still found.
2633 This behaviour is, however, controlled by an option called host_lookup_
2634 order, which defaults to "bydns:byaddr". If people want to use the other
2635 order, or indeed, just use one or the other means of lookup, they can
2636 specify it in this variable.
2638 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2639 some operating systems, this comes back from gethostbyaddr() as an empty
2640 string, and this is what Exim used to test for. However, it seems that in
2641 other systems, "." is yielded. Exim now tests for this case too.
2643 7. The values of check_spool_space and check_log_space are now held internally
2644 as a number of kilobytes instead of an absolute number of bytes. If a
2645 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2646 kilobyte. This means that much larger values can be stored.
2648 8. Exim monitor: an attempt to get the action menu when not actually pointing
2649 at a message produces an empty menu entitled "No message selected". This
2650 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2651 no entries in it ("Shell widget menu has zero width and/or height"). So I
2652 have added a single, blank menu entry in this case.
2654 9. Added ${quote_local_part.
265610. MIME decoding is now applied to the contents of Subject: header lines when
2657 they are logged.
265911. Now that a reference to $sender_host_address automatically causes a reverse
2660 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2661 host lookup before query-style lookups in lists that might use this
2662 variable. This has therefore been abolished, and the "net-" prefix is no
2663 longer necessary for query-style lookups.
266512. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2666 have been a typo for LFLAGS, so it has been changed.
266813. The install script calls Exim with "-C /dev/null" in order to find the
2669 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2670 to be output. Howeve, since Exim outputs its version number before the
2671 error, it didn't break the script. It just looked ugly. I fixed this by
2672 always allowing "-C /dev/null" if the caller is root.
267414. Ignore overlarge ACL variable number when reading spool file - insurance
2675 against a later release with more variables having written the file.
267715. The standard form for an IPv6 address literal was being rejected by EHLO.
2678 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2679 this, as well as the form without the "IPv6" on the front.
268116. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2682 OS/Makefile-Darwin file.
268417. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2685 only to LDAP libraries that do not have LDAP_OPT_DEREF.
268718. After change 4.21/52, "%ld" was used to format the contents of the $inode
2688 variable. However, some OS use ints for inodes. I've added cast to long int
2689 to get rid of the compiler warning.
269119. I had forgotten to lock out "/../" in configuration file names when
2692 ALT_CONFIG_PREFIX was set.
269420. Routers used for verification do not need to specify transports. However,
2695 if such a router generated a host list, and callout was configured, Exim
2696 crashed, because it could not find a port number from the (non-existent)
2697 transport. It now assumes port 25 in this circumstance.
269921. Added the -t option to exigrep.
270122. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2702 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2703 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2704 them is set, LOOKUP_LSEARCH is forced.
270623. "exim -bV" now outputs a list of lookups that are included in the binary.
270824. Added sender and host information to the "rejected by local_scan()" log
2709 line; previously there was no indication of these.
271125. Added .include_if_exists.
271326. Change 3.952/11 added an explicit directory sync on top of a file sync for
2714 Linux. It turns out that not all file systems support this. Apparently some
2715 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2716 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2717 not supported on the file descriptor, is now ignored when Exim is trying to
2718 sync a directory. This applies only to Linux.
272027. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
272228. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2723 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2724 this.
272629. Added support for maildirsize files from supplied patch (modified a bit).
272830. The use of :fail: followed by an empty string could lead Exim to respond to
2729 sender verification failures with (e.g.):
2731 550 Verification failed for <xxx>
2732 550 Sender verify failed
2734 where the first response line was missing the '-' that indicates it is not
2735 the final line of the response.
273731. The loop for finding the name of the user that called Exim had a hardwired
2738 limit of 10; it now uses the value of finduser_retries, which is used for
2739 all other user lookups.
274132. Added $received_count variable, available in data and not_smtp ACLs, and at
2742 delivery time.
274433. Exim was neglecting to zero errno before one call of strtol() when
2745 expanding a string and expecting an integer value. On some systems this
2746 resulted in spurious "integer overflow" errors. Also, it was casting the
2747 result into an int without checking.
274934. Testing for a connection timeout using "timeout_connect" in the retry rules
2750 did not work. The code looks as if it has *never* worked, though it appears
2751 to have been documented since at least releast 1.62. I have made it work.
275335. The "timeout_DNS" error in retry rules, also documented since at least
2754 1.62, also never worked. As it isn't clear exactly what this means, and
2755 clearly it isn't a major issue, I have abolished the feature by treating it
2756 as "timeout", and writing a warning to the main and panic logs.
275836. The display of retry rules for -brt wasn't always showing the error code
2759 correctly.
276137. Added new error conditions to retry rules: timeout_A, timeout_MX,
2762 timeout_connect_A, timeout_connect_MX.
276438. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2765 to the empty sender.
276739. The daemon was not analysing the content of -oX till after it had closed
2768 stderr and disconnected from the controlling terminal. This meant that any
2769 syntax errors were only noted on the panic log, and the return code from
2770 the command was 0. By re-arranging the code a little, I've made the
2771 decoding happen first, so such errors now appear on stderr, and the return
2772 code is 1. However, the actual setting up of the sockets still happens in
2773 the disconnected process, so errors there are still only recorded on the
2774 panic log.
277640. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2777 connections (as happens on some IP stacks) was logged at start up time as
2778 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2779 it from "IPv6 and IPv4", which means that two separate sockets are being
2780 used.
278241. The debug output for gethostbyname2() or getipnodebyname() failures now
2783 says whether AF_INET or AF_INET6 was passed as an argument.
278542. Exiwhat output was messed up when time zones were included in log
2786 timestamps.
278843. Exiwhat now gives more information about the daemon's listening ports,
2789 and whether -tls-on-connect was used.
279144. The "port" option of the smtp transport is now expanded.
279345. A "message" modifier in a "warn" statement in a non-message ACL was being
2794 silently ignored. Now an error message is written to the main and panic
2795 logs.
279746. There's a new ACL modifier called "logwrite" which writes to a log file
2798 as soon as it is encountered.
280047. Added $local_user_uid and $local_user_gid at routing time.
280248. Exim crashed when trying to verify a sender address that was being
2803 rewritten to "<>".
280549. Exim was recognizing only a space character after ".include". It now also
2806 recognizes a tab character.
280850. Fixed several bugs in the Perl script that creates the exim.8 man page by
2809 extracting the relevant information from the specification. The man page no
2810 longer contains scrambled data for the -d option, and I've added a section
2811 at the front about calling Exim under different names.
281351. Added "extra_headers" argument to the "mail" command in filter files.
281552. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2816 crash.
281853. Installed eximstats 1.29.
282054. Added transport_filter_timeout as a generic transport option.
282255. Exim no longer adds an empty Bcc: header to messages that have no To: or
2823 Cc: header lines. This was required by RFC 822, but it not required by RFC
2824 2822.
282656. Exim used to add From:, Date:, and Message-Id: header lines to any
2827 incoming messages that did not have them. Now it does so only if the
2828 message originates locally, that is, if there is no associated remote host
2829 address. When Resent- header lines are present, this applies to the Resent-
2830 lines rather than the non-Resent- lines.
283257. Drop incoming SMTP connection after too many syntax or protocol errors. The
2833 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
283558. Messages for configuration errors now include the name of the main
2836 configuration file - useful now that there may be more than one file in a
2837 list (.included file names were always shown).
283959. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2840 for those rare installations that do not start the daemon as root or run it
2841 setuid root. I've cut out the call to initgroups() if the daemon is not
2842 root at that time.
284460. The Exim user and group can now be bound into the binary as text strings
2845 that are looked up at the start of Exim's processing.
284761. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
284962. Added $mailstore_basename variable.
285163. Installed patch to sieve.c from Michael Haardt.
285364. When Exim failed to open the panic log after failing to open the main log,
2854 the original message it was trying to log was written to stderr and debug
2855 output, but if they were not available (the usual case in production), it
2856 was lost. Now it is written to syslog before the two lines that record the
2857 failures to open the logs.
285965. Users' Exim filters run in subprocesses under the user's uid. It is
2860 possible for a "deliver" command or an alias in a "personal" command to
2861 provoke an address rewrite. If logging of address rewriting is configured,
2862 this fails because the process is not running as root or exim. There may be
2863 a better way of dealing with this, but for the moment (because 4.30 needs
2864 to be released), I have disabled address rewrite logging when running a
2865 filter in a non-root, non-exim process.
2868Exim version 4.24
2871 1. The buildconfig auxiliary program wasn't quoting the value set for
2872 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2873 not defined. This bug was masked in 4.22 by the effect that was fixed in
2874 change 4.23/1.
2876 2. Some messages that were rejected after a message id was allocated were
2877 shown as "incomplete" by exigrep. It no longer does this for messages that
2878 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2880 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2881 have allow_domain_literals set, the ID did not get logged in the <= line.
2882 Domain literals are now always recognized in Message-ID: header lines.
2884 4. The first argument for a ${extract expansion item is the key name or field
2885 number. Leading and trailing spaces in this item were not being ignored,
2886 causing some misleading effects.
2888 5. When deliver_drop_privilege was set, single queue runner processes started
2889 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2890 same command in the process it spins off) were not dropping privilege.
2892 6. When the daemon running as "exim" started a queue runner, it always
2893 re-executed Exim in the spun-off process. This is a waste of effort when
2894 deliver_drop_privilege is set. The new process now just calls the
2895 queue-runner function directly.
2898Exim version 4.23
2901 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2904 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2905 ignored. Though the use of -oP was forcing the writing of a pid file, it
2906 was always written to the default place.
2908 3. If the message "no IP address found for host xxxx" is generated during
2909 incoming verification, it is now followed by identification of the incoming
2910 connection (so you can more easily find what provoked it).
2912 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2914 5. Added some features to "harden" Exim a bit more against certain attacks:
2916 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2917 be put in Local/Makefile. This is like the never_users runtime option,
2918 but it cannot be overridden. The default setting is "root".
2920 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2921 prefix string with which any file named in a -C command line option
2922 must start.
2924 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2925 is retained for -C and -D only if the caller of Exim is root. Without
2926 it, the exim user may also use -C and -D and retain privilege.
2928 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2929 command line option is disabled.
2931 6. Macro names set by the -D option must start with an upper case letter, just
2932 like macro names defined in the configuration file.
2934 7. Added "dereference=" facility to LDAP.
2936 8. Two instances of the typo "uknown" in the source files are fixed.
2938 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2939 the Configure-Makefile script screwed up while processing it.
294110. Incorporated PCRE 4.4.
294311. The SMTP synchronization check was not operating right at the start of an
2944 SMTP session. For example, it could not catch a HELO sent before the client
2945 waited for the greeting. There is now a check for outstanding input at the
2946 point when the greeting is written. Because of the duplex, asynchronous
2947 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2948 way, but not yet received, when the check is performed.
295012. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2951 on TCP/IP sockets, because this apparently causes some broken clients to
2952 timeout.
295413. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2955 unchanged) from the Cygwin maintainer.
295714. The code for -bV that shows what is in the binary showed "mbx" when maildir
2958 was supported instead of testing for mbx. Effectively a typo.
296015. The spa authenticator server code was not checking that the input it
2961 received was valid base64.
296316. The debug output line for the "set" modifier in ACLs was not showing the
2964 name of the variable that was being set.
296617. Code tidy: the variable type "vtype_string" was never used. Removed it.
296818. Previously, a reference to $sender_host_name did not cause a DNS reverse
2969 lookup on its own. Something else was needed to trigger the lookup. For
2970 example, a match in host_lookup or the need for a host name in a host list.
2971 Now, if $sender_host_name is referenced and the host name has not yet been
2972 looked up, a lookup is performed. If the lookup fails, the variable remains
2973 empty, and $host_lookup_failed is set to "1".
297519. Added "eqi" as a case-independent comparison operator.
297720. The saslauthd authentication condition could segfault if neither service
2978 nor realm was specified.
298021. If an overflowing value such as "2048M" was set for message_size_limit, the
2981 error message that was logged was misleading, and incoming SMTP
2982 connections were dropped. The message is now more accurate, and temporary
2983 errors are given to SMTP connections.
298522. In some error situations (such as 21 above) Exim rejects all SMTP commands
2986 (except RSET) with a 421 error, until QUIT is received. However, it was
2987 failing to send a response to QUIT.
298923. The HELO ACL was being run before the code for helo_try_verify_hosts,
2990 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2991 ACL is now run after the helo_try_verify_hosts code.
299324. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2994 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2995 case-independent, so other case variants are also recognized). Apparently
2996 some systems use these upper case variants.
299825. If more than two messages were waiting for the same host, and a transport
2999 filter was specified for the transport, Exim sent two messages over the
3000 same TCP/IP connection, and then failed with "socket operation on non-
3001 socket" when it tried to send the third.
300326. Added Exim::debug_write and Exim::log_write for embedded Perl use.
300527. The extern definition of crypt16() in expand.c was not being excluded when
3006 the OS had its own crypt16() function.
300828. Added bounce_return_body as a new option, and bounce_return_size_limit
3009 as a preferred synonym for return_size_limit, both as an option and as an
3010 expansion variable.
301229. Added LIBS=-liconv to OS/Makefile-OSF1.
301430. Changed the default configuration ACL to relax the local part checking rule
3015 for addresses that are not in any local domains. For these addresses,
3016 slashes and pipe symbols are allowed within local parts, but the sequence
3017 /../ is explicitly forbidden.
301931. SPA server authentication was not clearing the challenge buffer before
3020 using it.
302232. log_message in a "warn" ACL statement was writing to the reject log as
3023 well as to the main log, which contradicts the documentation and doesn't
3024 seem right (because no rejection is happening). So I have stopped it.
302633. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3027 However, I am unable to do any testing of this.
302934. Fixed an infelicity in the appendfile transport. When checking directories
3030 for a mailbox, to see if any needed to be created, it was accidentally
3031 using path names with one or more superfluous leading slashes; tracing
3032 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
303435. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3035 contents are added to the log line that is written for every discarded
3036 recipient. (Previously a log_message setting was ignored.)
303836. The ${quote: operator now quotes the string if it is empty.
304037. The install script runs exim in order to find its version number. If for
3041 some reason other than non-existence or emptiness, which it checks, it
3042 could not run './exim', it was installing it with an empty version number,
3043 i.e. as "exim-". This error state is now caught, and the installation is
3044 aborted.
304638. An argument was missing from the function that creates an error message
3047 when Exim fails to connect to the socket for saslauthd authentication.
3048 This could cause Exim to crash, or give a corrupted message.
305039. Added isip, isip4, and isip6 to ${if conditions.
305240. The ACL variables $acl_xx are now saved with the message, and can be
3053 accessed later in routers, transports, and filters.
305541. The new lookup type nwildlsearch is like wildlsearch, except that the key
3056 strings in the file are not string-expanded.
305842. If a MAIL command specified a SIZE value that was too large to fit into an
3059 int variable, the check against message_size_limit failed. Such values are
3060 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3061 day this will have to be increased, but I don't think I want to be around
3062 when emails are that large.
3066Exim version 4.22
3069 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3070 iconv() is not standard in FreeBSD.
3072 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3073 IPv6 enabled. The observed symptom was a segmentation fault on return from
3074 the function os_common_find_running_interfaces() in src/os.c.
3076 3. In the check_special_case() function in daemon.c I had used "errno" as an
3077 argument name, which causes warnings on some systems. This was basically a
3078 typo, since it was named "eno" in the comments!
3080 4. The code that waits for the clock to tick (at a resolution of some fraction
3081 of a second) so as to ensure message-id uniqueness was always waiting for
3082 at least one whole tick, when it could have waited for less. [This is
3083 almost certainly not relevant at current processor speeds, where it is
3084 unlikely to ever wait at all. But we try to future-proof.]
3086 5. The function that sleeps for a time interval that includes fractions of a
3087 second contained a race. It did not block SIGALRM between setting the
3088 timer, and suspending (a couple of lines later). If the interval was short
3089 and the sigsuspend() was delayed until after it had expired, the suspension
3090 never ended. On busy systems this could lead to processes getting stuck for
3091 ever.
3093 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3094 process, before it forks any delivery processes. The open lookup caching
3095 mechanism meant that the open file or database connection was passed into
3096 the delivery process. The problem was that delivery processes always tidy
3097 up cached lookup data. This could cause a problem for the next delivery
3098 process started by the queue runner, because the external queue runner
3099 process does not know about the closure. So the next delivery process
3100 still has data in the lookup cache. In the case of a file lookup, there was
3101 no problem because closing a file descriptor in a subprocess doesn't affect
3102 the parent. However, if the lookup was caching a connection to a database,
3103 the connection was closed, and the second delivery process was likely to
3104 see errors such as "PGSQL: query failed: server closed the connection
3105 unexpectedly". The problem has been fixed by closing all cached lookups
3106 in a queue runner before running a delivery process.
3108 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3109 seem to have the "const" qualifier which it has on other OS. I've
3110 parameterised it.
3112 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3113 *of the same type* (client or server) with the same public name that an
3114 error should be diagnosed.
3116 9. When Exim looked up a host name for an IP address, but failed to find the
3117 original IP address when looking up the host name (a safety check), it
3118 output the message "<ip address> does not match any IP for NULL", which was
3119 confusing, to say the least. The bug was that the host name should have
3120 appeared instead of "NULL".
312210. Since release 3.03, if Exim is called by a uid other than root or the Exim
3123 user that is built into the binary, and the -C or -D options is used, root
3124 privilege is dropped before the configuration file is read. In addition,
3125 logging is switched to stderr instead of the normal log files. If the
3126 configuration then re-defines the Exim user, the unprivileged environment
3127 is probably not what is expected, so Exim logs a panic warning message (but
3128 proceeds).
3130 However, if deliver_drop_privilege is set, the unprivileged state may well
3131 be exactly what is intended, so the warning has been cut out in that case,
3132 and Exim is allowed to try to write to its normal log files.
3135Exim version 4.21
3138 1. smtp_return_error_details was not giving details for temporary sender
3139 or receiver verification errors.
3141 2. Diagnose a configuration error if two authenticators have the same public
3142 name.
3144 3. Exim used not to create the message log file for a message until the first
3145 delivery attempt. This could be confusing when incoming messages were held
3146 for policy or load reasons. The message log file is now created at the time
3147 the message is received, and an initial "Received" line is written to it.
3149 4. The automatically generated man page for command line options had a minor
3150 bug that caused no ill effects; however, a more serious problem was that
3151 the procedure for building the man page automatically didn't always
3152 operate. Consequently, release 4.20 contains an out-of-date version. This
3153 shouldn't happen again.
3155 5. When building Exim with embedded Perl support, the script that builds the
3156 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3157 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3159 6. The freeze_tell option was not being used for messages that were frozen on
3160 arrival, either by an ACL or by local_scan().
3162 7. Added the smtp_incomplete_transaction log selector.
3164 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3165 was accepting AUTH without a new EHLO.
3167 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3168 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3169 RFC.
317110. Logging of TCP/IP connections (when configured) now happens in the main
3172 daemon process instead of the child process, so that the TCP/IP connection
3173 count is more accurate (but it can never be perfect).
317511. The use of "drop" in a nested ACL was not being handled correctly in the
3176 outer ACL. Now, if condition failure induced by the nested "drop" causes
3177 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3178 or "require"), the connection is dropped.
318012. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3181 that yield "discard" can now be used with an "accept" or a "discard" verb,
3182 but an error is generated for any others (because I can't see a useful way
3183 to define what should happen).
318513. When an ACL is read dynamically from a file (or anywhere else), the lines
3186 are now processed in the same way as lines in the Exim configuration file.
3187 In particular, continuation lines are supported.
318914. Added the "dnslists = a.b.c!=n.n.n.n" feature.
319115. Added -ti meaning -t -i.
319316. Check for letters, digits, hyphens, and dots in the names of dnslist
3194 domains, and warn by logging if others are found.
319617. At least on BSD, alignment is not guarenteed for the array of ifreq's
3197 returned from GIFCONF when Exim is trying to find the list of interfaces on
3198 a host. The code in os.c has been modified to copy each ifreq to an aligned
3199 structure in all cases.
3201 Also, in some cases, the returned ifreq's were being copied to a 'struct
3202 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3203 means the last couple of bytes of an IPv6 address could be chopped if the
3204 ifreq contained only a normal sockaddr (14 bytes storage).
320618. Named domain lists were not supported in the hosts_treat_as_local option.
3207 An entry such as +xxxx was not recognized, and was treated as a literal
3208 domain name.
321019. Ensure that header lines added by a DATA ACL are included in the reject log
3211 if the ACL subsequently rejects the message.
321320. Upgrade the utility script to use Digest::MD5 instead of just
3214 MD5 (which is deprecated).
321621. When testing a filter file using -bf, Exim was writing a message when it
3217 took the sender from a "From " line in the message, but it was not doing so
3218 when it took $return_path from a Return-Path: header line. It now does.
322022. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3221 with a valid header line field name (a series of printing characters
3222 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
322423. Changed "disc" in the source to "disk" to conform to the documentation and
3225 the book and for uniformity.
322724. Ignore Sendmail's -Ooption=value command line item.
322925. When execve() failed while trying to run a command in a pipe transport,
3230 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3231 could be confused with a return value of 69 from the command itself. This
3232 has been changed to 127, the value the shell returns if it is asked to run
3233 a non-existent command. The wording for the related log line suggests a
3234 non-existent command as the problem.
323626. If received_header_text expands to an empty string, do not add a Received:
3237 header line to the message. (Well, it adds a token one on the spool, but
3238 marks it "old" so that it doesn't get used or transmitted.)
324027. Installed eximstats 1.28 (addition of -nt option).
324228. There was no check for failure on the call to getsockname() in the daemon
3243 code. This can fail if there is a shortage of resources on the system, with
3244 ENOMEM, for example. A temporary error is now given on failure.
324629. Contrary to the C standard, it seems that in some environments, the
3247 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3248 program. Exim now does this explicitly; it affects the formatting of
3249 timestamps using strftime().
325130. If exiqsumm was given junk data, it threw up some uninitialized variable
3252 complaints. I've now initialized all the variables, to avoid this.
325432. Header lines added by a system filter were not being "seen" during
3255 transport-time rewrites.
325733. The info_callback() function passed to OpenSSL is set up with type void
3258 (*)(SSL *, int, int), as described somewhere. However, when calling the
3259 function (actually a macro) that sets it up, the type void(*)() is
3260 expected. I've put in a cast to prevent warnings from picky compilers.
326234. If a DNS black list lookup found a CNAME record, but there were no A
3263 records associated with the domain it pointed at, Exim crashed.
326535. If a DNS black list lookup returned more than one A record, Exim ignored
3266 all but the first. It now scans all returned addresses if a particular IP
3267 value is being sought. In this situation, the contents of the
3268 $dnslist_value variable are a list of all the addresses, separated by a
3269 comma and a space.
327136. Tightened up the rules for host name lookups using reverse DNS. Exim used
3272 to accept a host name and all its aliases if the forward lookup for any of
3273 them yielded the IP address of the incoming connection. Now it accepts only
3274 those names whose forward lookup yields the correct IP address. Any other
3275 names are discarded. This closes a loophole whereby a rogue DNS
3276 administrator could create reverse DNS records to break through a
3277 wildcarded host restriction in an ACL.
327937. If a user filter or a system filter that ran in a subprocess used any of
3280 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3281 the wrong values were passed to the pipe command ($thisaddress had the
3282 value of $0, $0 had the value of $1, etc). This bug was introduced by
3283 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
328538. Improved the line breaking for long SMTP error messages from ACLs.
3286 Previously, if there was no break point between 40 and 75 characters, Exim
3287 left the rest of the message alone. Two changes have been made: (a) I've
3288 reduced the minimum length to 35 characters; (b) if it can't find a break
3289 point between 35 and 75 characters, it looks ahead and uses the first one
3290 that it finds. This may give the occasional overlong line, but at least the
3291 remaining text gets split now.
329339. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3294 file descriptors might be low, and that setting 1000 would always raise it.
3295 It turns out that in some environments, the limit is already over 1000 and
3296 that lowering it causes trouble. So now Exim takes care not to decrease it.
329840. When delivering a message, the value of $return_path is set to $sender_
3299 address at the start of routing (routers may change the value). By an
3300 oversight, this default was not being set up when an address was tested by
3301 -bt or -bv, which affected the outcome if any router or filter referred to
3302 $return_path.
330441. The idea of the "warn" ACL verb is that it adds a header or writes to the
3305 log only when "message" or "log_message" are set. However, if one of the
3306 conditions was an address verification, or a call to a nested ACL, the
3307 messages generated by the underlying test were being passed through. This
3308 no longer happens. The underlying message is available in $acl_verify_
3309 message for both "message" and "log_message" expansions, so it can be
3310 passed through if needed.
331242. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3313 new expansion $bh_ to give the encoded byte string without charset
3314 translation. Translation happens only if iconv() is available; HAVE_ICONV
3315 indicates this at build time. HEADERS_CHARSET gives the charset to
3316 translate to; headers_charset can change it in the configuration, and
3317 "headers charset" can change it in an individual filter file.
331943. Now that we have a default RFC 2047 charset (see above), the code in Exim
3320 that creates RFC 2047 encoded "words" labels them as that charset instead
3321 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3322 expansion operator; (ii) when Exim creates a From: line for a local
3323 message; (iii) when a header line is rewritten to include a "phrase" part.
332544. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3326 buggy, causing it to skip the first lines of messages whose message ID
3327 ended in 'D'. This would not have bitten before Exim release 4.14, because
3328 message IDs were unlikely to end in 'D' before then. The effect was to have
3329 incorrect size information for certain domains.
333145. #include "config.h" was missing at the start of the crypt16.c module. This
3332 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3333 noticed.
333546. If there was a timeout during a "random" callout check, Exim treated it as
3336 a failure of the random address, and carried on sending RSET and the real
3337 address. If the delay was just some slowness somewhere, the response to the
3338 original RCPT would be taken as a response to RSET and so on, causing
3339 mayhem of various kinds.
334147. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3342 when I implemented it. It didn't allow for the fact that some option values
3343 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3344 the right test for positive values.
334648. Domain names in DNS records are case-independent. Exim always looks them up
3347 in lower case. Some resolvers return domain names in exactly the case they
3348 appear in the zone file, that is, they may contain uppercase letters. Not
3349 all resolvers do this - some return always lower case. Exim was treating a
3350 change of case by a resolver as a change of domain, similar to a widening
3351 of a domain abbreviation. This triggered its re-routing code and so it was
3352 trying to route what was effectively the same domain again. This normally
3353 caused routing to fail (because the router wouldn't handle the domain
3354 twice). Now Exim checks for this case specially, and just changes the
3355 casing of the domain that it ultimately uses when it transmits the message
3356 envelope.
335849. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3359 module.
336150. If a filter generated a file delivery with a non-absolute name (possible if
3362 no home directory exists for the router), the forbid_file option was not
3363 forbidding it.
336551. Added '&' feature to dnslists, to provide bit mask matching in addition to
3366 the existing equality matching.
336852. Exim was using ints instead of ino_t variables in some places where it was
3369 dealing with inode numbers.
337153. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3372 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3373 called TMPDIR, and if it finds it is different, it changes its value.
337554. The smtp_printf() function is now made available to local_scan() so
3376 additional output lines can be written before returning. There is also an
3377 smtp_fflush() function to enable the detection of a dropped connection.
3378 The variables smtp_input and smtp_batched_input are exported to
3379 local_scan().
338155. Changed the default runtime configuration: the message "Unknown user"
3382 has been removed from the ACL, and instead placed on the localuser router,
3383 using the cannot_route_message feature. This means that any verification
3384 failures that generate their own messages won't get overridden. Similarly,
3385 the "Unrouteable address" message that was in the ACL for unverifiable
3386 relay addresses has also been removed.
338856. Added hosts_avoid_esmtp to the smtp transport.
339057. The exicyclog script was not checking for the esoteric option
3391 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3392 will work only if exicyclog is run under the appropriate euid.
339458. Following a discussion on the list, the rules by which Exim recognises line
3395 endings on incoming messages have been changed. The -dropcr and drop_cr
3396 options are now no-ops, retained only for backwards compatibility. The
3397 following line terminators are recognized: LF CRLF CR. However, special
3398 processing applies to CR:
3400 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3401 nor a local message in the state where . is a terminator.
3403 (ii) If a bare CR is encountered in a header line, an extra space is added
3404 after the line terminator so as not to end the header. The reasoning
3405 behind this is that bare CRs in header lines are most likely either
3406 to be mistakes, or people trying to play silly games.
340859. The size of a message, as listed by "-bp" or in the Exim monitor window,
3409 was being incorrectly given as 18 bytes larger than it should have been.
3410 This is a VOB (very old bug).
341260. This may never have affected anything current, but just in case it has:
3413 When the local host is found other than at the start of a list of hosts,
3414 the local host, those with the same MX, and any that follow, are discarded.
3415 When the list in question was part of a longer list of hosts, the following
3416 hosts (not currently being processed) were also being discarded. This no
3417 longer happens. I'm not sure if this situation could ever has previously
3418 arisen.
342061. Added the "/MX" feature to lists of hosts in the manualroute and query
3421 program routers.
342362. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3424 header. This is something that is recommended in a new Internet Draft, and
3425 is something that is documented as being done by Sendmail. There are two
3426 possible values. For messages generated by the autoreply transport, Exim
3427 adds:
3429 Auto-Submitted: auto-replied
3431 whereas for all other generated messages (e.g. bounces) it adds
3433 Auto-Submitted: auto-generated
343563. The "personal" condition in filters now includes a test for the
3436 Auto-Submitted: header. If it contains the string "auto-" the message it
3437 not considered personal.
343964. Added rcpt_include_affixes as a generic transport option.
344165. Added queue_only_override (default true).
344366. Added the syslog_duplication option.
344567. If what should have been the first header line of a message consisted of
3446 a space followed by a colon, Exim was mis-interpreting it as a header line.
3447 It isn't of course - it is syntactically invalid and should therefore be
3448 treated as the start of the message body. The misbehaviour could have
3449 caused a number of strange effects, including loss of data in subsequent
3450 header lines, and spool format errors.
345268. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3453 client host had authenticated. This control can now be exercised by an ACL
3454 for more flexibility.
345669. By default, callouts do not happen when testing with -bh. There is now a
3457 variant, -bhc, which does actually run the callout code, including
3458 consulting and updating the callout cache.
346070. Added support for saslauthd authentication, courtesy of Alexander
3461 Sabourenkov.
346371. If statvfs() failed on the spool or log directories while checking their
3464 size for availability, Exim confusingly gave the error "space shortage".
3465 Furthermore, in debugging mode it crashed with a floating point exception.
3466 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3467 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3468 serious problem, Exim now writes to the main and panic logs when this
3469 happens, with details of the failure. It then refuses to accept the
3470 incoming message, giving the message "spool directory problem" or "log
3471 directory problem" with a 421 code for SMTP messages.
347372. When Exim is about to re-exec itself, it ensures that the file descriptors
3474 0, 1, and 2 exist, because some OS complain for execs without them (see
3475 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3476 descriptors. However, the code omitted to check that the open succeeded,
3477 causing mysterious errors if for some reason the permissions on /dev/null
3478 got screwed. Now Exim writes a message to the main and panic logs, and
3479 bombs out if it can't open /dev/null.
348173. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3482 interact so that it is all more flexible. It is supposed to remain
3483 backwards compatible. Also added extra_local_interfaces.
348574. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3486 to bomb out with an assertion failure - to the client this appears as a
3487 connection drop. This problem occurs in the part of the code that was taken
3488 from the Samba project. Fortunately, the assertion is in a very simple
3489 function, so I have fixed this by reproducing the function inline in the
3490 one place where it is called, and arranging for authentication to fail
3491 instead of killing the process with assert().
349375. The SPA client code was not working when the server requested OEM rather
3494 than Unicode encoding.
349676. Added code to make require_files with a specific uid setting more usable in
3497 the case where statting the file as root fails - usually a non-root-mounted
3498 NFS file system. When this happens and the failure is EACCES, Exim now
3499 forks a subprocess and does the per-uid checking as the relevant uid.
350177. Added process_log_path.
350378. If log_file_path was not explicitly set, a setting of check_log_space or
3504 check_log_inodes was ignored.
350679. If a space check for the spool or log partitions fails, the incident is now
3507 logged. Of course, in the latter case the data may get lost...
350980. Added the %p formatting code to string_format() so that it can be used to
3510 print addresses in debug_print(). Adjusted all the address printing in the
3511 debugging in store.c to use %p rather than %d.
351381. There was a concern that a line of code in smtp_in.c could overflow a
3514 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3515 initially expressed, the concern was not well-founded, because trailing
3516 spaces are removed early. However, if the trailing spaces were followed by
3517 a NULL, they did not get removed, so the overflow was possible. Two fixes
3518 were applied:
3520 (a) I re-wrote the offending code in a cleaner fashion.
3521 (b) If an incoming SMTP command contains a NULL character, it is rejected
3522 as invalid.
352482. When Exim changes uid/gid to the Exim user at daemon start time, it now
3525 runs initgroups(), so that if the Exim user is in any additional groups,
3526 they will be used during message reception.
3529Exim version 4.20
3532The change log for 4.20 and earlier releases has been archived.