1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.240 2005/09/26 09:52:18 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Added the "success_on_redirect" address verification option. See
10 NewStuff for rationale and an example.
12 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
14 PH/02 Patch to exigrep to allow it to work on syslog lines.
16 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
17 fread() to skip over the body file's header line, because in Cygwin the
18 header line is locked and is inaccessible.
20 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
21 co-exist for some time) to make it clear that it is the Exim ID that is
22 referenced, not the Message-ID: header line.
24 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
25 string_format() function, because snprintf() does not exist on all
28 PH/06 The use of forbid_filter_existstest now also locks out the use of the
29 ${stat: expansion item.
31 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
32 protocol synchronization error", to keep the pedants happy.
34 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
35 well as for IRIX systems, when gcc is being used. See the host.c source
38 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
40 PH/10 Named domain lists were not working if used in a queue_smtp_domains
43 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
44 transport and to the smtp transport in LMTP mode.
46 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
48 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
49 run a filter in a subprocess. This could lead to confusion in subsequent
50 lookups in the parent process. There should also be a search_tidyup() at
51 the end of the subprocess.
53 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
54 only if the host matched helo_try_verify_hosts, which caused the
55 verification to occur when the EHLO/HELO command was issued. The ACL just
56 tested the remembered result. Now, if a previous verification attempt has
57 not happened, "verify = helo" does it there and then.
59 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
61 TK/03 Fix log output including CR from clamd.
63 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
64 exist provoked a memory error which could cause a segfault.
66 PH/15 Installed PCRE 6.2
68 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
70 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
71 of the problem. Specifically, suggested +O2 rather than +O1 for the
74 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
76 PH/20 If a delivery was routed to a non-standard port by means of an SRV
77 record, the port was not correctly logged when the outgoing_port log
78 selector was set (it logged the transort's default port).
80 PH/21 Added support for host-specific ports to manualroute, queryprogram,
81 fallback_hosts, and "hosts" in the smtp transport.
83 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
84 host errors such as "Connection refused".
86 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
87 authentication with radiusclient 0.4.9:
89 - Error returned from rc_read_config was caught wrongly
90 - Username/password not passed on to radius server due to wrong length.
92 The presumption is that some radiusclient API changes for 4.51/PH/17
93 were not taken care of correctly. The code is still untested by me (my
94 Linux distribution still has 0.3.2 of radiusclient), but it was
95 contributed by a Radius user.
97 PH/24 When doing a callout, the value of $domain wasn't set correctly when
98 expanding the "port" option of the smtp transport.
100 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
101 while reading a MIME header. Thanks to Tom Hughes for a patch.
103 PH/24 Include config.h inside local_scan.h so that configuration settings are
106 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
107 that in an ACL for RCPT (for example), you can examine exactly what was
110 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
111 commands, but it was not correctly comparing the address with the actual
112 client host address. Thus, it would show the EHLO address in Received:
113 header lines when this was not necessary.
115 PH/27 Added the % operator to ${eval:}.
117 PH/28 Exim tries to create and chdir to its spool directory when it starts;
118 it should be ignoring failures (because with -C, for example, it has lost
119 privilege). It wasn't ignoring creation failures other than "already
122 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
125 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
126 because an input line was too long, either on its own, or by virtue of
127 too many continuations, the temporary file was not being removed, and the
128 return code was incorrect.
130 PH/31 Missing "BOOL" in function definition in filtertest.c.
132 PH/32 Applied Sieve patches from the maintainer.
134 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
136 PH/33 Added "verify = not_blind".
138 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
139 Local/Makefile (with some defaults set). These are used in built scripts
140 such as exicyclog, but they have never been used in the exim_install
141 script (though there are many overriding facilities there). I have
142 arranged that the exim_install script now takes note of these two
145 PH/35 Installed configuration files for Dragonfly.
147 PH/36 When a locally submitted message by a trusted user did not contain a
148 From: header, and the sender address was obtained from -f or from an SMTP
149 MAIL command, and the trusted user did not use -F to supply a sender
150 name, $originator_name was incorrectly used when constructing a From:
151 header. Furthermore, $originator_name was used for submission mode
152 messages from external hosts without From: headers in a similar way,
153 which is clearly wrong.
155 PH/37 Added control=suppress_local_fixups.
157 PH/38 When log_selector = +received_sender was set, and the addition of the
158 sender made the log line's construction buffer exactly full, or one byte
159 less than full, an overflow happened when the terminating "\n" was
162 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
163 when the result of a list match is failure because a DNS lookup failed.
165 PH/40 RM_COMMAND is now used in the building process.
167 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
168 the "build-* directories that it finds.
170 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
171 address in a domain literal was a prefix of an interface address.
173 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
174 when verifying a sender address, unless rewrite_headers is false.
176 PH/44 Wrote a long comment about why errors_to addresses are verified as
177 recipients, not senders.
179 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
180 the ratelimit ACL was added.
182 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
184 PH/46 Added extra information about PostgreSQL errors to the error string.
186 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
187 avoiding the need to install special zones in a real server. This is
188 backwards compatible; if it can't find the fake resolver, it drops back.
189 Thus, both old and new test suites can be run.
191 TF/02 Added util/ratelimit.pl
193 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
194 clock is set back in time.
196 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
197 Candler <B.Candler@pobox.com>.
199 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
200 OK for addresses that are the result of redirections.
202 PH/48 A number of further additions for the benefit of the new test suite,
203 including a fake gethostbyname() that interfaces to the fake DNS resolver
206 TF/06 The fix for widen_domains has also been applied to qualify_single and
207 search_parents which are the other dnslookup options that can cause
210 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
213 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
215 TF/07 Exim produced the error message "an SRV record indicated no SMTP
216 service" if it encountered an MX record with an empty target hostname.
217 The message is now "an MX or SRV record indicated no SMTP service".
219 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
220 if the DNS of the sending site is misconfigured. This is quite a
221 common situation. This change restores the behaviour of treating a
222 helo verification defer as a failure.
224 PH/51 If self=fail was set on a router, the bounce message did not include the
225 actual error message.
231 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
233 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
234 command itself ended in error, the underlying message about the transport
235 filter timeout was being overwritten with the pipe command error. Now the
236 underlying error message should be appended to the second error message.
238 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
240 PH/02 Reduce the amount of output that "make" produces by default. Full output
241 can still be requested.
243 PH/03 The warning log line about a condition test deferring for a "warn" verb
244 was being output only once per connection, rather than after each
245 occurrence (because it was using the same function as for successful
246 "warn" verbs). This seems wrong, so I have changed it.
248 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
249 it should not have, which might have caused a crash in the right
250 circumstances, but probably never did.
252 PH/04 Installed a modified version of Tony Finch's patch to make submission
253 mode fix the return path as well as the Sender: header line, and to
254 add a /name= option so that you can make the user's friendly name appear
257 TF/03 Added the control = fakedefer ACL modifier.
259 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
260 Mark Lowes for thorough testing.
262 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
264 TK/03 Merged latest SRS patch from Miles Wilton.
266 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
267 with the definition in sysexits.h (which is #included earlier).
268 Fortunately, Exim does not actually use EX_OK. The code used to try to
269 preserve the sysexits.h value, by assumimg that macro definitions were
270 scanned for macro replacements. I have been disabused of this notion,
271 so now the code just undefines EX_OK before #including unistd.h.
273 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
274 in the smtp transport. When a block could not be written in a single
275 write() function, the timeout was being re-applied to each part-write.
276 This seems wrong - if the receiver was accepting one byte at a time it
277 would take for ever. The timeout is now adjusted when this happens. It
278 doesn't have to be particularly precise.
280 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
281 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
283 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
284 without a domain if the check to <postmaster@domain> fails.
286 SC/01 Eximstats: added -xls and the ability to specify output files
287 (patch written by Frank Heydlauf).
289 SC/02 Eximstats: use FileHandles for outputing results.
291 SC/03 Eximstats: allow any combination of xls, txt, and html output.
293 SC/04 Eximstats: fixed display of large numbers with -nvr option
295 SC/05 Eximstats: fixed merging of reports with empty tables.
297 SC/06 Eximstats: added the -include_original_destination flag
299 SC/07 Eximstats: removed tabs and trailing whitespace.
301 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
303 TK/06 MBOX spool code: Add real "From " MBOX separator line
304 so the .eml file is really in mbox format (even though
305 most programs do not really care). Patch from Alex Miller.
307 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
308 The latter is generated from $received_to and is only set if the
309 message has one envelope recipient. SA can use these headers,
310 obviously out-of-the-box. Patch from Alex Miller.
312 PH/08 The ${def test on a variable was returning false if the variable's
313 value was "0", contrary to what the specification has always said!
314 The result should be true unless the variable is empty.
316 PH/09 The syntax error of a character other than { following "${if
317 def:variable_name" (after optional whitespace) was not being diagnosed.
318 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
319 accidental colon was present, for example, could give incorrect results.
321 PH/10 Tidied the code in a number of places where the st_size field of a stat()
322 result is used (not including appendfile, where other changes are about
325 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
326 This involved changing a lot of size variables from int to off_t. It
327 should work with maildirs and everything.
329 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
330 spamd dying while we are connected to it.
332 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
335 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
336 function for BDB 4.3.
338 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
340 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
341 the sources for PCRE have been re-arranged and more files are now
344 PH/15 The code I had for printing potentially long long variables in PH/11
345 above was not the best (it lost precision). The length of off_t variables
346 is now inspected at build time, and an appropriate printing format (%ld
347 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
348 to be "long long int" or "long int". This is needed for the internal
349 formatting function string_vformat().
351 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
352 the configuration file to be ":syslog", then the script "guesses" where
353 the logs files are, rather than using the compiled in default. In our
354 case the guess is not the same as the compiled default, so the script
355 suddenly stopped working when I started to use syslog. The patch checks
356 to see if log_file_path is "". If so, it attempts to read it from exim
357 with no configuration file to get the compiled in version, before it
358 falls back to the previous guessing code."
360 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
361 implementing BATV in an Exim configuration. See NewStuff for the gory
364 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
365 Makefile that are specific to HP-UX.
367 PH/18 If the "use_postmaster" option was set for a recipient callout together
368 with the "random" option, the postmaster address was used as the MAIL
369 FROM address for the random test, but not for the subsequent recipient
370 test. It is now used for both.
372 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
373 patch removes a few documentation additions to RFC 3028, because the
374 latest draft now contains them. It adds the new en;ascii-case comparator
375 and a new error check for 8bit text in MIME parts. Comparator and
376 require names are now matched exactly. I enabled the subaddress
377 extension, but it is not well tested yet (read: it works for me)."
379 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
380 rework some of the code of TK/09 above to avoid the hardwired use of
381 "%lld" and "long long". Replaced the call to snprintf() with a call to
384 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
385 records point to non-existent hosts", "retry timeout exceeded", and
386 "retry time not reached for any host after a long failure period".
388 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
389 experimental DomainKeys support:
391 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
392 (2) On an error such as an illegally used "control", the wrong name for
393 the control was given.
395 These problems did NOT occur unless DomainKeys support was compiled.
397 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
399 PH/24 Added ${if match_ip condition.
401 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
402 databases so that it will be absolutely obvious if a crash occurs in the
403 DB library. This is a regular occurrence (often caused by mis-matched
406 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
407 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
408 picked up on a user's system that detects such things. There doesn't seem
409 to be a gcc warning option for this - only an attribute that has to be
410 put on the function's prototype. It seems that in Fedora Core 4 they have
411 set this on a number of new functions. No doubt there will be more in due
414 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
415 specify a transport. However, if an address that was verified by such a
416 router was the subject of a callout, Exim crashed because it tried to
417 read the rcpt_include_affixes from the non-existent transport. Now it
418 just assumes that the setting of that option is false. This bug was
419 introduced by 4.51/PH/31.
421 PH/28 Changed -d+all to exclude +memory, because that information is very
422 rarely of interest, but it makes the output a lot bigger. People tend to
423 do -d+all out of habit.
425 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
426 code in os-type was giving problems when libc.so lives in lib64, like on
429 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
430 aren't the modern standard, and it seems that some systems' include files
431 don't always have them. Exim was already checking for some of the newer
432 ones like T_AAAA, and defining it itself. I've added checks for all the
433 record types that Exim uses.
435 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
436 not automatically generating a new one, as it is supposed to. This
437 prevented TLS from working. If the file did exist, but contained invalid
438 data, a new version was generated, as expected. It was only the case of a
439 non-existent file that was broken.
441 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
442 with a change in libdomainkeys > 0.64.
444 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
445 from DNS. If the selector record carries the flag, it now has
446 precedence over the domain-wide flag.
448 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
450 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
451 the use of a port name that isn't defined in /etc/services) occurred, the
452 message was deferred as in a normal delivery, and thus remained on the
453 spool, instead of being failed because of the mua_wrapper setting. This
454 is now fixed, and I tidied up some of the mua_wrapper messages at the
457 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
458 the messages in a hash of arrays rather than using individual hashes.
459 This is a bit cleaner and results in dramatic memory savings, albeit
460 at a slight CPU cost.
462 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
463 as requested by Marc Sherman.
465 SC/10 Eximstats: added histograms for user specified patterns as requested
468 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
470 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
471 fopen() in the content-scanning modules that did not already have it.
477 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
478 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
480 TK/02 Fix ACL "control" statement not being available in MIME ACL.
482 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
484 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
485 to test Sieve filters that use "vacation".
487 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
488 that changes the way the GnuTLS parameters are stored in the cache file.
489 The new format can be generated externally. For backward compatibility,
490 if the data in the cache doesn't make sense, Exim assumes it has read an
491 old-format file, and it generates new data and writes a new file. This
492 means that you can't go back to an older release without removing the
495 PH/03 A redirect router that has both "unseen" and "one_time" set does not
496 work if there are any delivery delays because "one_time" forces the
497 parent to be marked "delivered", so its unseen clone is never tried
498 again. For this reason, Exim now forbids the simultaneous setting of
501 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
502 redirected to themselves ("homonym" addresses). Read the long ChangeLog
503 entry if you want to know the details. The fix, however, neglected to
504 consider the case when local delivery batching is involved. The test for
505 "previously delivered" was not happening when checking to see if an
506 address could be batched with a previous (undelivered) one; under
507 certain circumstances this could lead to multiple deliveries to the same
510 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
511 in its include files, and this causes problems building Exim.
513 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
514 header_syntax) but Exim was just ignoring anything given after a slash.
515 In particular, this caused confusion with an attempt to use "verify =
516 reverse_host_lookup/defer_ok". An error is now given when options are
517 supplied for verify items that do not have them. (Maybe reverse_host_
518 lookup should have a defer_ok option, but that's a different point.)
520 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
521 defined by RFC 821) to 2048, because there were problems with some AUTH
522 commands, and RFC 1869 says the size should be increased for extended
523 SMTP commands that take arguments.
525 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
528 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
529 "unknown" error; now it says that the functionality isn't in the binary.
531 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
532 an address' error message when a string expansion fails (syntax or
533 whatever). Otherwise the password may appear in the log. Following change
534 PH/42 below, there is no longer a chance of it appearing in a bounce
537 PH/11 Installed exipick version 20050225.0 from John Jetmore.
539 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
540 of its addresses was ever tried. (Bugzilla bug #2.)
542 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
543 the result incorrectly in the debug output. (It correctly added a newline
544 to what was transported.)
546 TF/01 Added $received_time.
548 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
549 commented out examples of how to interface to a virus scanner and to
550 SpamAssassin. Also added commented examples of av_scanner and
551 spamd_address settings.
553 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
554 and controls are allowed in which ACLs. There were a couple of minor
555 errors. Some of the entries in the conditions table (which is a table of
556 where they are NOT allowed) were getting very unwieldy; rewrote them as a
557 negation of where the condition IS allowed.
559 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
561 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
562 header file does not have a version number, so I've had to invent a new
563 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
564 API. The code is untested by me (my Linux distribution still has 0.3.2 of
565 radiusclient), but it was contributed by a Radius user.
567 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
568 files or directories, for OpenSSL.
570 PH/19 When an Exim process that is running as root has to create an Exim log
571 file, it does so in a subprocess that runs as exim:exim so as to get the
572 ownership right at creation (otherwise, other Exim processes might see
573 the file with the wrong ownership). There was no test for failure of this
574 fork() call, which would lead to the process getting stuck as it waited
575 for a non-existent subprocess. Forks do occasionally fail when resources
576 run out. I reviewed all the other calls to fork(); they all seem to check
579 PH/20 When checking for unexpected SMTP input at connect time (before writing
580 the banner), Exim was not dealing correctly with a non-positive return
581 from the read() function. If the client had disconnected by this time,
582 the result was a log entry for a synchronization error with an empty
583 string after "input=" when read() returned zero. If read() returned -1
584 (an event I could not check), uninitialized data bytes were printed.
585 There were reports of junk text (parts of files, etc) appearing after
588 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
590 PH/22 Added support for macro redefinition, and (re)definition in between
591 driver and ACL definitions.
593 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
594 forgetting to use the resulting value; it was using the unexpanded value.
596 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
597 hadn't been configured. The fix is from Juergen Kreileder, who
598 understands it better than I do:
600 "Here's what I see happening with three configured cyrus_sasl
601 authenticators configured (plain, login, cram-md5):
603 On startup auth_cyrus_sasl_init() gets called for each of these.
604 This means three calls to sasl_listmech() without a specified mech_list.
605 => SASL tests which mechs of all available mechs actually work
606 => three warnings about OTP not working
607 => the returned list contains: plain, login, cram-md5, digest-md5, ...
609 With the patch, sasl_listmech() also gets called three times. But now
610 SASL's mech_list option is set to the server_mech specified in the the
611 authenticator. Or in other words, the answer from sasl_listmech()
612 gets limited to just the mech you're testing for (which is different
614 => the return list contains just 'plain' or 'login', 'cram-md5' or
615 nothing depending on the value of ob->server_mech.
617 I've just tested the patch: Authentication still works fine,
618 unavailable mechs specified in the exim configuration are still
619 caught, and the auth.log warnings about OTP are gone."
621 PH/25 When debugging is enabled, the contents of the command line are added
622 to the debugging output, even when log_selector=+arguments is not
625 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
626 answer is "GNU", and only if the return is "GNU/something" is the answer
629 PH/27 $acl_verify_message is now set immediately after the failure of a
630 verification in an ACL, and so is available in subsequent modifiers. In
631 particular, the message can be preserved by coding like this:
633 warn !verify = sender
634 set acl_m0 = $acl_verify_message
636 Previously, $acl_verify_message was set only while expanding "message"
637 and "log_message" when a very denied access.
639 PH/28 Modified OS/os.c-Linux with
641 -#ifndef OS_LOAD_AVERAGE
642 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
644 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
645 nomenclature these days.)
647 PH/29 Installed patch from the Sieve maintainer that adds the options
648 sieve_useraddress and sieve_subaddress to the redirect router.
650 PH/30 In these circumstances:
651 . Two addresses routed to the same list of hosts;
652 . First host does not offer TLS;
653 . First host accepts first address;
654 . First host gives temporary error to second address;
655 . Second host offers TLS and a TLS session is established;
656 . Second host accepts second address.
657 Exim incorrectly logged both deliveries with the TLS parameters (cipher
658 and peerdn, if requested) that were in fact used only for the second
661 PH/31 When doing a callout as part of verifying an address, Exim was not paying
662 attention to any local part prefix or suffix that was matched by the
663 router that accepted the address. It now behaves in the same way as it
664 does for delivery: the affixes are removed from the local part unless
665 rcpt_include_affixes is set on the transport.
667 PH/32 Add the sender address, as F=<...>, to the log line when logging a
668 timeout during the DATA phase of an incoming message.
670 PH/33 Sieve envelope tests were broken for match types other than :is. I have
671 applied a patch sanctioned by the Sieve maintainer.
673 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
674 the uid or gid is negative. A case of a negative gid caused this to be
675 noticed. The fix allows for either to be negative.
677 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
678 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
681 PH/36 The change PH/12 above was broken. Fixed it.
683 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
684 the grounds that routing the same address twice would always produce the
685 same answer. This might have been true once, but it is certainly no
686 longer true now. Routing a child address may depend on the previous
687 routing that produced that child. Some complicated redirection strategies
688 went wrong when messages had multiple recipients, and made Exim's
689 behaviour dependent on the order in which the addresses were given.
691 I have moved the duplicate checking until after the routing is complete.
692 Exim scans the addresses that are assigned to local and remote
693 transports, and removes any duplicates. This means that more work will be
694 done, as duplicates will always all be routed, but duplicates are
695 presumably rare, so I don't expect this is of any significance.
697 For deliveries to pipes, files, and autoreplies, the duplicate checking
698 still happens during the routing process, since they are not going to be
701 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
702 It corrects a timeout issue with spamd. This is Ian's comment: "The
703 background is that sometimes spamd either never reads data from a
704 connection it has accepted, or it never writes response data. The exiscan
705 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
706 blindly assumes that writes won't block so it may never time out."
708 PH/39 Allow G after quota size as well as K and M.
710 PH/40 The value set for $authenticated_id in an authenticator may not contain
711 binary zeroes or newlines because the value is written to log lines and
712 to spool files. There was no check on this. Now the value is run through
713 the string_printing() function so that such characters are converted to
714 printable escape sequences.
716 PH/41 $message_linecount is a new variable that contains the total number of
717 lines in the message. Compare $body_linecount, which is the count for the
720 PH/42 Exim no longer gives details of delivery errors for specific addresses in
721 bounce and delay warning messages, except in certain special cases, which
724 (a) An SMTP error message from a remote host;
725 (b) A message specified in a :fail: redirection;
726 (c) A message specified in a "fail" command in a system filter;
727 (d) A message specified in a FAIL return from the queryprogram router;
728 (e) A message specified by the cannot_route_message router option.
730 In these cases only, Exim does include the error details in bounce and
731 warning messages. There are also a few cases where bland messages such
732 as "unrouteable address" or "local delivery error" are given.
734 PH/43 $value is now also set for the "else" part of a ${run expansion.
736 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
737 being worked on, but at least Exim now implements the latest version to
740 PH/45 In a pipe transport, although a timeout while waiting for the pipe
741 process to complete was treated as a delivery failure, a timeout while
742 writing the message to the pipe was logged, but erroneously treated as a
743 successful delivery. Such timeouts include transport filter timeouts. For
744 consistency with the overall process timeout, these timeouts are now
745 treated as errors, giving rise to delivery failures by default. However,
746 there is now a new Boolean option for the pipe transport called
747 timeout_defer, which, if set TRUE, converts the failures into defers for
748 both kinds of timeout. A transport filter timeout is now identified in
751 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
752 systems where "make" and "gmake" are different, calling "gmake" at top
753 level broke things. I've arranged for the value of $(MAKE) to be passed
754 from the Makefile to this script so that it can call the same version of
758 A note about Exim versions 4.44 and 4.50
759 ----------------------------------------
761 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
762 changes of various kinds. As a consequence, a big documentation update was
763 needed. This delayed the release for rather longer than seemed good, especially
764 in the light of a couple of (minor) security issues. Therefore, the changes
765 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
766 release. So 4.44 and 4.50 are in effect two different branches that both start
769 I have left the 4.50 change log unchanged; it contains all the changes since
770 4.43. The change log for 4.44 is below; many of its items are identical to
771 those for 4.50. This seems to be the most sensible way to preserve the
772 historical information.
778 1. Minor wording change to the doc/README.SIEVE file.
780 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
781 computation of the current number of files was incorrect.
783 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
784 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
787 4. Give more explanation in the error message when the command for a transport
788 filter fails to execute.
790 5. There are several places where Exim runs a non-Exim command in a
791 subprocess. The SIGUSR1 signal should be disabled for these processes. This
792 was being done only for the command run by the queryprogram router. It is
793 now done for all such subprocesses. The other cases are: ${run, transport
794 filters, and the commands run by the lmtp and pipe transports.
796 6. Added CONFIGURE_GROUP build-time option.
798 7. Some older OS have a limit of 256 on the maximum number of file
799 descriptors. Exim was using setrlimit() to set 1000 as a large value
800 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
801 systems. I've change it so that if it can't get 1000, it tries for 256.
803 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
804 was an oversight, and furthermore, ever since the addition of extra
805 controls (e.g. 4.43/32), the checks on when to allow different forms of
806 "control" were broken. There should now be diagnostics for all cases when a
807 control that does not make sense is encountered.
809 9. Added the /retain_sender option to "control=submission".
811 10. $recipients is now available in the predata ACL (oversight).
813 11. Tidy the search cache before the fork to do a delivery from a message
814 received from the command line. Otherwise the child will trigger a lookup
815 failure and thereby defer the delivery if it tries to use (for example) a
816 cached ldap connection that the parent has called unbind on.
818 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
819 of $address_data from the recipient verification was clobbered by the
822 13. The value of address_data from a sender verification is now available in
823 $sender_address_data in subsequent conditions in the ACL statement.
825 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
827 15. Added a new option "connect=<time>" to callout options, to set a different
830 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
831 was its contents. (It was OK if the option was not defined at all.)
833 17. A "Completed" log line is now written for messages that are removed from
834 the spool by the -Mrm option.
836 18. New variables $sender_verify_failure and $recipient_verify_failure contain
837 information about exactly what failed.
839 19. Added -dd to debug only the daemon process.
841 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
842 handles timeouts, both on the server side and network timeouts. Renamed the
843 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
845 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
846 It is now set to "smtps".
848 22. $host_address is now set to the target address during the checking of
851 23. When checking ignore_target_hosts for an ipliteral router, no host name was
852 being passed; this would have caused $sender_host_name to have been used if
853 matching the list had actually called for a host name (not very likely,
854 since this list is usually IP addresses). A host name is now passed as
857 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
858 code that specifies a non-restarting handler (typically sigaction() in
859 modern systems) in an attempt to fix a rare and obscure crash bug.
861 25. Narrowed the window for a race in the daemon that could cause it to ignore
862 SIGCHLD signals. This is not a major problem, because they are used only to
863 wake it up if nothing else does.
865 26. A malformed maildirsize file could cause Exim to calculate negative values
866 for the mailbox size or file count. Odd effects could occur as a result.
867 The maildirsize information is now recalculated if the size or filecount
870 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
871 support for a long time. Removed HAVE_SYS_VFS_H.
873 28. Installed the latest version of exipick from John Jetmore.
875 29. In an address list, if the pattern was not a regular expression, an empty
876 subject address (from a bounce message) matched only if the pattern was an
877 empty string. Non-empty patterns were not even tested. This was the wrong
878 because it is perfectly reasonable to use an empty address as part of a
879 database query. An empty address is now tested by patterns that are
880 lookups. However, all the other forms of pattern expect the subject to
881 contain a local part and a domain, and therefore, for them, an empty
882 address still always fails if the pattern is not itself empty.
884 30. Exim went into a mad DNS loop when attempting to do a callout where the
885 host was specified on an smtp transport, and looking it up yielded more
888 31. Re-factored the code for checking spool and log partition space into a
889 function that finds that data and another that does the check. The former
890 is then used to implement four new variables: $spool_space, $log_space,
891 $spool_inodes, and $log_inodes.
893 32. The RFC2047 encoding function was originally intended for short strings
894 such as real names; it was not keeping to the 75-character limit for
895 encoded words that the RFC imposes. It now respects the limit, and
896 generates multiple encoded words if necessary. To be on the safe side, I
897 have increased the buffer size for the ${rfc2047: expansion operator from
900 33. It is now permitted to omit both strings after an "if" condition; if the
901 condition is true, the result is "true". As before, when the second string
902 is omitted, a false condition yields an empty string. This makes it less
903 cumbersome to write custom ACL and router conditions.
905 34. Failure to deliver a bounce message always caused it to be frozen, even if
906 there was an errors_to setting on the router. The errors_to setting is now
909 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
910 canonical form (fully expanded) before being placed in
911 $sender_host_address.
913 36. The table in the code that translates DNS record types into text (T_A to
914 "A" for instance) was missing entries for NS and CNAME. It is just possible
915 that this could have caused confusion if both these types were looked up
916 for the same domain, because the text type is used as part of Exim's
917 per-process caching. But the chance of anyone hitting this buglet seems
920 37. The dnsdb lookup has been extended in a number of ways.
922 (1) There is a new type, "zns", which walks up the domain tree until it
923 finds some nameserver records. It should be used with care.
925 (2) There is a new type, "mxh", which is like "mx" except that it returns
926 just the host names, not the priorities.
928 (3) It is now possible to give a list of domains (or IP addresses) to be
929 looked up. The behaviour when one of the lookups defers can be
930 controlled by a keyword.
932 (4) It is now possible to specify the separator character for use when
933 multiple records are returned.
935 38. The dnslists ACL condition has been extended: it is now possible to supply
936 a list of IP addresses and/or domains to be looked up in a particular DNS
939 39. Added log_selector=+queue_time_overall.
941 40. When running the queue in the test harness, wait just a tad after forking a
942 delivery process, to get repeatability of debugging output.
944 41. Include certificate and key file names in error message when GnuTLS fails
945 to set them up, because the GnuTLS error message doesn't include the name
946 of the failing file when there is a problem reading it.
948 42. Allow both -bf and -bF in the same test run.
950 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
952 44. The "Exiscan patch" is now merged into the mainline Exim source.
954 45. Sometimes the final signoff response after QUIT could fail to get
955 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
956 < 0 before doing a fflush(). This bug looks as though it goes back to the
957 introduction of TLS in release 3.20, but "sometimes" must have been rare
958 because the tests only now provoked it.
960 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
961 (this can affect the format of dates).
963 47. exim_tidydb, when checking for the continued existence of a message for
964 which it has found a message-specific retry record, was not finding
965 messages that were in split spool directories. Consequently, it was
966 deleting retry records that should have stayed in existence.
968 48. Steve fixed some bugs in eximstats.
970 49. The SPA authentication driver was not abandoning authentication and moving
971 on to the next authenticator when an expansion was forced to fail,
972 contradicting the general specification for all authenticators. Instead it
973 was generating a temporary error. It now behaves as specified.
975 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
976 (the order specifies the preference for clients). The order is now AES256,
977 AES128, 3DES, ARCFOUR128.
979 51. Small patch to Sieve code - explicitly set From: when generating an
982 52. Exim crashed if a remote delivery caused a very long error message to be
983 recorded - for instance if somebody sent an entire SpamAssassin report back
984 as a large number of 550 error lines. This bug was coincidentally fixed by
985 increasing the size of one of Exim's internal buffers (big_buffer) that
986 happened as part of the Exiscan merge. However, to be on the safe side, I
987 have made the code more robust (and fixed the comments that describe what
990 53. Now that there can be additional text after "Completed" in log lines (if
991 the queue_time_overall log selector is set), a one-byte patch to exigrep
992 was needed to allow it to recognize "Completed" as not the last thing in
995 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
996 patch that reportedly fixes this has been added. I am not expert enough to
997 create a test for it. This is what the patch creator wrote:
999 "I found a little strange behaviour of ldap code when working with
1000 Windows 2003 AD Domain, where users was placed in more than one
1001 Organization Units. When I tried to give exim partial DN, the exit code
1002 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1003 But simultaneously result of request was absolutely normal ldap result,
1004 so I produce this patch..."
1006 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1007 REFERENCE, so I have modified the code to exclude the patch when that macro
1010 55. Some experimental protocols are using DNS PTR records for new purposes. The
1011 keys for these records are domain names, not reversed IP addresses. The
1012 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1013 leaves it alone. Component reversal etc. now happens only for IP addresses.
1016 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1018 57. Double the size of the debug message buffer (to 2048) so that more of very
1019 long debug lines gets shown.
1021 58. The exicyclog utility now does better if the number of log files to keep
1022 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1024 59. Two changes related to the smtp_active_hostname option:
1026 (1) $smtp_active_hostname is now available as a variable.
1027 (2) The default for smtp_banner uses $smtp_active_hostname instead
1028 of $primary_hostname.
1030 60. The host_aton() function is supposed to be passed a string that is known
1031 to be a valid IP address. However, in the case of IPv6 addresses, it was
1032 not checking this. This is a hostage to fortune. Exim now panics and dies
1033 if the condition is not met. A case was found where this could be provoked
1034 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1035 components; fortuitously, this particular loophole had already been fixed
1036 by change 4.50/55 above.
1038 If there are any other similar loopholes, the new check in host_aton()
1039 itself should stop them being exploited. The report I received stated that
1040 data on the command line could provoke the exploit when Exim was running as
1041 exim, but did not say which command line option was involved. All I could
1042 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1043 running as the user.
1046 61. There was a buffer overflow vulnerability in the SPA authentication code
1047 (which came originally from the Samba project). I have added a test to the
1048 spa_base64_to_bits() function which I hope fixes it.
1051 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1052 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1054 63. The daemon start-up calls getloadavg() while still root for those OS that
1055 need the first call to be done as root, but it missed one case: when
1056 deliver_queue_load_max is set with deliver_drop_privilege. This is
1057 necessary for the benefit of the queue runner, because there is no re-exec
1058 when deliver_drop_privilege is set.
1060 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1061 This has been fixed.
1063 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1064 was in use, was not putting the data itself into the right store pool;
1065 consequently, it could be overwritten for a subsequent message in the same
1066 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1069 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1071 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1072 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1073 the function were treating the return as a boolean value, which happened to
1074 work because 0=false and not-0=true, but is not correct code.
1076 68. The host_aton() function was not handling scoped IPv6 addresses (those
1077 with, for example, "%eth0" on the end) correctly.
1079 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1080 negated items (that is, ~something) in unsigned ints. Some compilers
1081 apparently mutter when there is no cast.
1083 70. If an address verification called from an ACL failed, and did not produce a
1084 user-specific message (i.e. there was only a "system" message), nothing was
1085 put in $acl_verify_message. In this situation, it now puts the system
1088 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1089 session; change 4.31/43 added the unwanted input to the log line - except
1090 that it did not do this in the start of session case. It now does.
1092 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1093 This is wrong and can cause the other end to generate a synchronization
1094 error if it is another Exim or anything else that does the synchronization
1095 check. A QUIT command is no longer sent after a timeout.
1097 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1098 during host lookups.
1100 74. The defer_ok option of callout verification was not working if it was used
1101 when verifying addresses in header lines, that is, for this case:
1103 verify = header_sender/callout=defer_ok
1105 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1106 those file descriptors could be used for SMTP connections. If anything
1107 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1108 could be sent to the SMTP client, causing chaos. The daemon now opens
1109 stdin, stdout, and stderr to /dev/null when it puts itself into the
1112 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1113 log by default. The user can override this with suitable Perl magic.
1115 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1116 the log message when discard triggers, was not working for the DATA ACL or
1117 for the non-SMTP ACL.
1119 78. Error message wording change in sieve.c.
1121 79. If smtp_accept_max_per_host was set, the number of connections could be
1122 restricted to fewer than expected, because the daemon was trying to set up
1123 a new connection before checking whether the processes handling previous
1124 connections had finished. The check for completed processes is now done
1125 earlier. On busy systems, this bug wouldn't be noticed because something
1126 else would have woken the daemon, and it would have reaped the completed
1129 80. If a message was submitted locally by a user whose login name contained one
1130 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1131 It caused a spool format error. I have fixed the spool reading code. A
1132 related problem was that the "from" clause in the Received: line became
1133 illegal because of the space(s). It is now covered by ${quote_local_part.
1135 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1138 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1140 83. Patch from Sieve maintainer to fix unterminated string problem in
1141 "vacation" handling.
1143 84. Some minor changes to the Linux configuration files to help with other
1144 OS variants using glibc.
1146 85. One more patch for Sieve to update vacation handling to latest spec.
1149 ----------------------------------------------------
1150 See the note above about the 4.44 and 4.50 releases.
1151 ----------------------------------------------------
1157 1. Change 4.43/35 introduced a bug that caused file counts to be
1158 incorrectly computed when quota_filecount was set in an appendfile
1161 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1162 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1165 3. Give more explanation in the error message when the command for a transport
1166 filter fails to execute.
1168 4. There are several places where Exim runs a non-Exim command in a
1169 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1170 was being done only for the command run by the queryprogram router. It is
1171 now done for all such subprocesses. The other cases are: ${run, transport
1172 filters, and the commands run by the lmtp and pipe transports.
1174 5. Some older OS have a limit of 256 on the maximum number of file
1175 descriptors. Exim was using setrlimit() to set 1000 as a large value
1176 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1177 systems. I've change it so that if it can't get 1000, it tries for 256.
1179 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1180 was an oversight, and furthermore, ever since the addition of extra
1181 controls (e.g. 4.43/32), the checks on when to allow different forms of
1182 "control" were broken. There should now be diagnostics for all cases when a
1183 control that does not make sense is encountered.
1185 7. $recipients is now available in the predata ACL (oversight).
1187 8. Tidy the search cache before the fork to do a delivery from a message
1188 received from the command line. Otherwise the child will trigger a lookup
1189 failure and thereby defer the delivery if it tries to use (for example) a
1190 cached ldap connection that the parent has called unbind on.
1192 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1193 of $address_data from the recipient verification was clobbered by the
1194 sender verification.
1196 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1197 was its contents. (It was OK if the option was not defined at all.)
1199 11. A "Completed" log line is now written for messages that are removed from
1200 the spool by the -Mrm option.
1202 12. $host_address is now set to the target address during the checking of
1203 ignore_target_hosts.
1205 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1206 being passed; this would have caused $sender_host_name to have been used if
1207 matching the list had actually called for a host name (not very likely,
1208 since this list is usually IP addresses). A host name is now passed as
1211 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1212 code that specifies a non-restarting handler (typically sigaction() in
1213 modern systems) in an attempt to fix a rare and obscure crash bug.
1215 15. Narrowed the window for a race in the daemon that could cause it to ignore
1216 SIGCHLD signals. This is not a major problem, because they are used only to
1217 wake it up if nothing else does.
1219 16. A malformed maildirsize file could cause Exim to calculate negative values
1220 for the mailbox size or file count. Odd effects could occur as a result.
1221 The maildirsize information is now recalculated if the size or filecount
1224 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1225 support for a long time. Removed HAVE_SYS_VFS_H.
1227 18. Updated exipick to current release from John Jetmore.
1229 19. Allow an empty sender to be matched against a lookup in an address list.
1230 Previously the only cases considered were a regular expression, or an
1233 20. Exim went into a mad DNS lookup loop when doing a callout where the
1234 host was specified on the transport, if the DNS lookup yielded more than
1237 21. The RFC2047 encoding function was originally intended for short strings
1238 such as real names; it was not keeping to the 75-character limit for
1239 encoded words that the RFC imposes. It now respects the limit, and
1240 generates multiple encoded words if necessary. To be on the safe side, I
1241 have increased the buffer size for the ${rfc2047: expansion operator from
1244 22. Failure to deliver a bounce message always caused it to be frozen, even if
1245 there was an errors_to setting on the router. The errors_to setting is now
1248 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1249 canonical form (fully expanded) before being placed in
1250 $sender_host_address.
1252 24. Updated eximstats to version 1.33
1254 25. Include certificate and key file names in error message when GnuTLS fails
1255 to set them up, because the GnuTLS error message doesn't include the name
1256 of the failing file when there is a problem reading it.
1258 26. Expand error message when OpenSSL has problems setting up cert/key files.
1261 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1262 (this can affect the format of dates).
1264 28. exim_tidydb, when checking for the continued existence of a message for
1265 which it has found a message-specific retry record, was not finding
1266 messages that were in split spool directories. Consequently, it was
1267 deleting retry records that should have stayed in existence.
1269 29. eximstats updated to version 1.35
1270 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1271 1.35 - bugfix such that pie charts by volume are generated correctly
1273 30. The SPA authentication driver was not abandoning authentication and moving
1274 on to the next authenticator when an expansion was forced to fail,
1275 contradicting the general specification for all authenticators. Instead it
1276 was generating a temporary error. It now behaves as specified.
1278 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1279 (the order specifies the preference for clients). The order is now AES256,
1280 AES128, 3DES, ARCFOUR128.
1282 31. Small patch to Sieve code - explicitly set From: when generating an
1285 32. Exim crashed if a remote delivery caused a very long error message to be
1286 recorded - for instance if somebody sent an entire SpamAssassin report back
1287 as a large number of 550 error lines. This bug was coincidentally fixed by
1288 increasing the size of one of Exim's internal buffers (big_buffer) that
1289 happened as part of the Exiscan merge. However, to be on the safe side, I
1290 have made the code more robust (and fixed the comments that describe what
1293 33. Some experimental protocols are using DNS PTR records for new purposes. The
1294 keys for these records are domain names, not reversed IP addresses. The
1295 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1296 leaves it alone. Component reversal etc. now happens only for IP addresses.
1299 34. The host_aton() function is supposed to be passed a string that is known
1300 to be a valid IP address. However, in the case of IPv6 addresses, it was
1301 not checking this. This is a hostage to fortune. Exim now panics and dies
1302 if the condition is not met. A case was found where this could be provoked
1303 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1304 components; fortuitously, this particular loophole had already been fixed
1305 by change 4.50/55 or 4.44/33 above.
1307 If there are any other similar loopholes, the new check in host_aton()
1308 itself should stop them being exploited. The report I received stated that
1309 data on the command line could provoke the exploit when Exim was running as
1310 exim, but did not say which command line option was involved. All I could
1311 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1312 running as the user.
1315 35. There was a buffer overflow vulnerability in the SPA authentication code
1316 (which came originally from the Samba project). I have added a test to the
1317 spa_base64_to_bits() function which I hope fixes it.
1320 36. The daemon start-up calls getloadavg() while still root for those OS that
1321 need the first call to be done as root, but it missed one case: when
1322 deliver_queue_load_max is set with deliver_drop_privilege. This is
1323 necessary for the benefit of the queue runner, because there is no re-exec
1324 when deliver_drop_privilege is set.
1326 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1327 was in use, was not putting the data itself into the right store pool;
1328 consequently, it could be overwritten for a subsequent message in the same
1329 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1332 38. Sometimes the final signoff response after QUIT could fail to get
1333 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1334 < 0 before doing a fflush(). This bug looks as though it goes back to the
1335 introduction of TLS in release 3.20, but "sometimes" must have been rare
1336 because the tests only now provoked it.
1342 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1343 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1344 Now it returns an integer. A number of calls were still expecting a T/F
1345 return. Fortuitously, in all cases, the tests worked in OK situations,
1346 which is the norm. However, things would have gone wrong on any write
1347 failures on the smtp file descriptor. This function is used when sending
1348 messages over SMTP and also when doing verify callouts.
1350 2. When Exim is called to do synchronous delivery of a locally submitted
1351 message (the -odf or -odi options), it no longer closes stderr before doing
1354 3. Implemented the mua_wrapper option.
1356 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1358 5. Implemented the functions header_remove(), header_testname(),
1359 header_add_at_position(), and receive_remove_recipient(), and exported them
1362 6. If an ACL "warn" statement specified the addition of headers, Exim already
1363 inserted X-ACL-Warn: at the start if there was no header name. However, it
1364 was not making this test for the second and subsequent header lines if
1365 there were newlines in the string. This meant that an invalid header could
1366 be inserted if Exim was badly configured.
1368 7. Allow an ACL "warn" statement to add header lines at the start or after all
1369 the Received: headers, as well as at the end.
1371 8. Added the rcpt_4xx retry error code.
1373 9. Added postmaster_mailfrom=xxx to callout verification option.
1375 10. Added mailfrom=xxxx to the callout verification option, for verify=
1378 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1379 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1381 12. Inserted some casts to stop certain compilers warning when using pointer
1382 differences as field lengths or precisions in printf-type calls (mostly
1383 affecting debugging statements).
1385 13. Added optional readline() support for -be (dynamically loaded).
1387 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1388 same clock tick as a message's arrival, so that its received time was the
1389 same as the "first fail" time on the retry record, and that message
1390 remained on the queue past the ultimate address timeout, every queue runner
1391 would try a delivery (because it was past the ultimate address timeout) but
1392 after another failure, the ultimate address timeout, which should have then
1393 bounced the address, did not kick in. This was a "< instead of <=" error;
1394 in most cases the first failure would have been in the next clock tick
1395 after the received time, and all would be well.
1397 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1398 being recognized when the domain list was tested by the match_domain
1399 condition in an expansion string.
1401 16. Added the ${str2b64: operator.
1403 17. Exim was always calling setrlimit() to set a large limit for the number of
1404 processes, without checking whether the existing limit was already
1405 adequate. (It did check for the limit on file descriptors.) Furthermore,
1406 errors from getrlimit() and setrlimit() were being ignored. Now they are
1407 logged to the main and panic logs, but Exim does carry on, to try to do its
1408 job under whatever limits there are.
1410 18. Imported PCRE 5.0.
1412 19. Trivial typo in log message " temporarily refused connection" (the leading
1415 20. If the log selector return_path_on_delivery was set and an address was
1416 redirected to /dev/null, the delivery process crashed because it assumed
1417 that a return path would always be set for a "successful" delivery. In this
1418 case, the whole delivery is bypassed as an optimization, and therefore no
1421 21. Internal re-arrangement: the function for sending a challenge and reading
1422 a response while authentication was assuming a zero-terminated challenge
1423 string. It's now changed to take a pointer and a length, to allow for
1424 binary data in such strings.
1426 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1428 23. Exim was not respecting finduser_retries when seeking the login of the
1429 uid under which it was called; it was always trying 10 times. (The default
1430 setting of finduser_retries is zero.) Also, it was sleeping after the final
1431 failure, which is pointless.
1433 24. Implemented tls_on_connect_ports.
1435 25. Implemented acl_smtp_predata.
1437 26. If the domain in control=submission is set empty, Exim assumes that the
1438 authenticated id is a complete email address when it generates From: or
1439 Sender: header lines.
1441 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1442 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1443 chown and chgrp in /bin and hostname in /usr/bin.
1445 28. Exim was keeping the "process log" file open after each use, just as it
1446 does for the main log. This opens the possibility of it remaining open for
1447 long periods when the USR1 signal hits a daemon. Occasional processlog
1448 errors were reported, that could have been caused by this. Anyway, it seems
1449 much more sensible not to leave this file open at all, so that is what now
1452 29. The long-running daemon process does not normally write to the log once it
1453 has entered its main loop, and it closes the log before doing so. This is
1454 so that log files can straightforwardly be renamed and moved. However,
1455 there are a couple of unusual error situations where the daemon does write
1456 log entries, and I had neglected to close the log afterwards.
1458 30. The text of an SMTP error response that was received during a remote
1459 delivery was being truncated at 512 bytes. This is too short for some of
1460 the long messages that one sometimes sees. I've increased the limit to
1463 31. It is now possible to make retry rules that apply only when a message has a
1464 specific sender, in particular, an empty sender.
1466 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1467 it possible to be selective about when SMTP synchronization is enforced.
1469 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1471 32. Implemented hosts_connection_nolog.
1473 33. Added an ACL for QUIT.
1475 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1478 35. Added mailbox_size and mailbox_filecount to appendfile.
1480 36. Added control = no_multiline_responses to ACLs.
1482 37. There was a bug in the logic of the code that waits for the clock to tick
1483 in the case where the clock went backwards by a substantial amount such
1484 that the microsecond fraction of "now" was more than the microsecond
1485 fraction of "then" (but the whole seconds number was less).
1487 38. Added support for the libradius Radius client library this is found on
1488 FreeBSD (previously only the radiusclient library was supported).
1494 1. When certain lookups returned multiple values in the form name=value, the
1495 quoting of the values was not always being done properly. Specifically:
1496 (a) If the value started with a double quote, but contained no whitespace,
1498 (b) If the value contained whitespace other than a space character (i.e.
1499 tabs or newlines or carriage returns) it was not quoted.
1500 This fix has been applied to the mysql and pgsql lookups by writing a
1501 separate quoting function and calling it from the lookup code. The fix
1502 should probably also be applied to nisplus, ibase and oracle lookups, but
1503 since I cannot test any of those, I have not disturbed their existing code.
1505 2. A hit in the callout cache for a specific address caused a log line with no
1506 reason for rejecting RCPT. Now it says "Previous (cached) callout
1507 verification failure".
1509 3. There was an off-by-one bug in the queryprogram router. An over-long
1510 return line was truncated at 256 instead of 255 characters, thereby
1511 overflowing its buffer with the terminating zero. As well as fixing this, I
1512 have increased the buffer size to 1024 (and made a note to document this).
1514 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1515 when Exim is waiting for an SMTP response from a remote server, Exim
1516 restarts its select() call on the socket, thereby resetting its timeout.
1517 This is not a problem when such interrupts are rare. Somebody set up a cron
1518 job to run exiwhat every 2 minutes, which is less than the normal select()
1519 timeout (5 or 10 minutes). This meant that the select() timeout never
1520 kicked in because it was always reset. I have fixed this by comparing the
1521 time when an interrupt arrives with the time at the start of the first call
1522 to select(). If more time than the timeout has elapsed, the interrupt is
1523 treated as a timeout.
1525 5. Some internal re-factoring in preparation for the addition of Sieve
1526 extensions (by MH). In particular, the "personal" test is moved to a
1527 separate function, and given an option for scanning Cc: and Bcc: (which is
1528 not set for Exim filters).
1530 6. When Exim created an email address using the login of the caller as the
1531 local part (e.g. when creating a From: or Sender: header line), it was not
1532 quoting the local part when it contained special characters such as @.
1534 7. Installed new OpenBSD configuration files.
1536 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1537 try to make them clearer.
1539 9. Callout options, other than the timeout value, were being ignored when
1540 verifying sender addresses in header lines. For example, when using
1542 verify = header_sender/callout=no_cache
1544 the cache was (incorrectly) being used.
1546 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1547 only the Cygwin environment.
1549 11. When return_path_on_delivery was set as a log selector, if different remote
1550 addresses in the same message used different return paths and parallel
1551 remote delivery occurred, the wrong values would sometimes be logged.
1552 (Whenever a remote delivery process finished, the return path value from
1553 the most recently started remote delivery process was logged.)
1555 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1556 lines when AUTH and/or TLS are in use. This is the "received protocol"
1557 field. Exim used to use "asmtp" for authenticated SMTP, without any
1558 indication (in the protocol name) for TLS use. Now it follows the RFC and
1559 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1560 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1561 names appear in log lines as well as in Received: header lines.
1563 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1564 extensions, and comparison tests, and to fix some bugs.
1566 14. Changes to the "personal" filter test:
1568 (1) The test was buggy in that it was just doing the equivalent of
1569 "contains" tests on header lines. For example, if a user's address was
1570 anne@some.where, the "personal" test would incorrectly be true for
1572 To: susanne@some.where
1574 This test is now done by extracting each address from the header in turn,
1575 and checking the entire address. Other tests that are part of "personal"
1576 are now done using regular expressions (for example, to check local parts
1577 of addresses in From: header lines).
1579 (2) The list of non-personal local parts in From: addresses has been
1580 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1581 taken from the Sieve specification recommendations.
1583 (3) If the message contains any header line starting with "List-" it is
1584 treated as non-personal.
1586 (4) The test for "circular" in the Subject: header line has been removed
1587 because it now seems ill-conceived.
1589 15. Minor typos in src/EDITME comments corrected.
1591 16. Installed latest exipick from John Jetmore.
1593 17. If headers_add on a router specified a text string that was too long for
1594 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1595 of string_sprintf() is now avoided.
1597 18. $message_body_size was not set (it was always zero) when running the DATA
1598 ACL and the local_scan() function.
1600 19. For the "mail" command in an Exim filter, no default was being set for
1601 the once_repeat time, causing a random time value to be used if "once" was
1602 specified. (If the value happened to be <= 0, no repeat happened.) The
1603 default is now 0s, meaning "never repeat". The "vacation" command was OK
1604 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1605 (I found it when inspecting the code).
1607 20. There is now an overall timeout for performing a callout verification. It
1608 defaults to 4 times the callout timeout, which applies to individual SMTP
1609 commands during the callout. The overall timeout applies when there is more
1610 than one host that can be tried. The timeout is checked before trying the
1611 next host. This prevents very long delays if there are a large number of
1612 hosts and all are timing out (e.g. when the network connections are timing
1613 out). The value of the overall timeout can be changed by specifying an
1614 additional sub-option for "callout", called "maxwait". For example:
1616 verify = sender/callout=5s,maxwait=20s
1618 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1619 to the end before writing, but this should make it even safer).
1621 22. Exim was forgetting that it had advertised PIPELINING for the second and
1622 subsequent messages on an SMTP connection. It was also not resetting its
1623 memory on STARTTLS and an internal HELO.
1625 23. When Exim logs an SMTP synchronization error within a session, it now
1626 records whether PIPELINING has been advertised or not.
1628 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1629 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1630 rather than long int.
1632 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1634 26. Added the never_mail option to autoreply.
1640 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1641 crash if the getsockname() call failed; this can happen if a connection is
1642 closed very soon after it is established. The problem was simply in the
1643 order in which certain operations were done, causing Exim to try to write
1644 to the SMTP stream before it had set up the file descriptor. The bug has
1645 been fixed by making things happen in the correct order.
1651 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1652 before the connection was closed, thus losing the rejection response.
1654 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1655 some early Solaris releases, but causes trouble in current releases where
1656 socklen_t is defined.
1658 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1661 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1664 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1665 that releasing the top of it at the end releases what was used for sub-
1666 expansions (unless the block got too big). However, discard this block if
1667 the first thing is a variable or header, so that we can use its block when
1668 it is dynamic (useful for very large $message_headers, for example).
1670 6. Lookups now cache *every* query, not just the most recent. A new, separate
1671 store pool is used for this. It can be recovered when all lookup caches are
1672 flushed. Lookups now release memory at the end of their result strings.
1673 This has involved some general refactoring of the lookup sources.
1675 7. Some code has been added to the store_xxx() functions to reduce the amount
1676 of flapping under certain conditions.
1678 8. log_incoming_interface used to affect only the <= reception log lines. Now
1679 it causes the local interface and port to be added to several more SMTP log
1680 lines, for example "SMTP connection from", and rejection lines.
1682 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1684 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1686 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1687 could be overwritten at the end of the current message (or the start of a
1688 new message if it was set in a HELO ACL). The value is now preserved for
1689 the duration of the SMTP connection.
1691 12. If a transport had a headers_rewrite setting, and a matching header line
1692 contained an unqualified address, that address was qualified, even if it
1693 did not match any rewriting rules. The underlying bug was that the values
1694 of the flags that permit the existence of unqualified sender and recipient
1695 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1696 non-local messages, and by -bnq for local messages) were not being
1697 preserved with the message after it was received.
1699 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1700 "next input=" as part of the text comprising the host identity instead of
1701 the correct text. The code was using the same buffer for two different
1702 strings. However, depending on which order the printing function evaluated
1703 its arguments, the bug did not always show up. Under Linux, for example, my
1704 test suite worked just fine.
1706 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1707 doesn't work with some older versions of Perl. It has been changed to "my",
1708 which in any case is probably the better facility to use.
1710 15. A really picky compiler found some instances of statements for creating
1711 error messages that either had too many or two few arguments for the format
1714 16. The size of the buffer for calls to the DNS resolver has been increased
1715 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1716 for addresses that have a lot of PTR records. This alleviates a problem; it
1717 does not fully solve it.
1719 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1720 buffer now truncates the list and logs the incident, which is the same
1721 action as happens when Exim is looking up a host name and its aliases.
1722 Previously in this situation something unpredictable would happen;
1723 sometimes it was "internal error: store_reset failed".
1725 18. If a server dropped the connection unexpectedly when an Exim client was
1726 using GnuTLS and trying to read a response, the client delivery process
1727 crashed while trying to generate an error log message.
1729 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1730 string, for example:
1732 warn message = H1: something\nH2: something
1734 the text was added as a single header line from Exim's point of view
1735 though it ended up OK in the delivered message. However, searching for the
1736 second and subsequent header lines using $h_h2: did not work. This has been
1737 fixed. Similarly, if a system filter added multiple headers in this way,
1738 the routers could not see them.
1740 20. Expanded the error message when iplsearch is called with an invalid key to
1741 suggest using net-iplsearch in a host list.
1743 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1744 ACLs are no longer actually imposed (and a message to that effect is
1747 22. If a "gecos" field in a passwd entry contained escaped characters, in
1748 particular, if it contained a \" sequence, Exim got it wrong when building
1749 a From: or a Sender: header from that name. A second bug also caused
1750 incorrect handling when an unquoted " was present following a character
1751 that needed quoting.
1753 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1754 was not being matched caselessly.
1756 24. Arranged for all hyphens in the exim.8 source to be escaped with
1759 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1760 properly. Recipient callout cache records were still being keyed to include
1761 the sender, even when use_sender was set false. This led to far more
1762 callouts that were necessary. The sender is no longer included in the key
1763 when use_sender is false.
1765 26. Added "control = submission" modifier to ACLs.
1767 27. Added the ${base62d: operator to decode base 62 numbers.
1769 28. dnsdb lookups can now access SRV records.
1771 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1772 the configuration file.
1774 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1775 (-v) mode. This makes the output for a verbose queue run more intelligible.
1777 31. Added a use_postmaster feature to recipient callouts.
1779 32. Added the $body_zerocount variable, containing the number of binary zero
1780 bytes in the message body.
1782 33. The time of last modification of the "new" subdirectory is now used as the
1783 "mailbox time last read" when there is a quota error for a maildir
1786 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1788 35. Added +ignore_unknown as a special item in host lists.
1790 36. Code for decoding IPv6 addresses in host lists is now included, even if
1791 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1792 address was recognized as an IP address, but was then not correctly decoded
1793 into binary, causing unexpected and incorrect effects when compared with
1800 1. Very minor rewording of debugging text in manualroute to say "list of
1801 hosts" instead of "hostlist".
1803 2. If verify=header_syntax was set, and a header line with an unqualified
1804 address (no domain) and a large number of spaces between the end of the
1805 name and the colon was received, the reception process suffered a buffer
1806 overflow, and (when I tested it) crashed. This was caused by some obsolete
1807 code that should have been removed. The fix is to remove it!
1809 3. When running in the test harness, delay a bit after writing a bounce
1810 message to get a bit more predictability in the log output.
1812 4. Added a call to search_tidyup() just before forking a reception process. In
1813 theory, someone could use a lookup in the expansion of smtp_accept_max_
1814 per_host which, without the tidyup, could leave open a database connection.
1816 5. Added the variables $recipient_data and $sender_data which get set from a
1817 lookup success in an ACL "recipients" or "senders" condition, or a router
1818 "senders" option, similar to $domain_data and $local_part_data.
1820 6. Moved the writing of debug_print from before to after the "senders" test
1823 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1824 problems for message scanning, either using a data ACL, or using
1825 local_scan() because the Received: header was not generated till after they
1826 were called (in order to set the time as the time of reception completion).
1827 I have revised the way this works. The header is now generated after the
1828 body is received, but before the ACL or local_scan() are called. After they
1829 are run, the timestamp in the header is updated.
1835 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1836 before starting a queue runner without re-exec. This happened only when
1837 deliver_drop_privilege was set or when the Exim user was set to root. The
1838 effect of the bug was that timeouts during subsequent deliveries caused
1839 crashes instead of being properly handled. The handler is now left at its
1840 default (and expected) setting.
1842 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1843 message, again when deliver_drop_privilege is set or Exim is run as root.
1844 The bug described in (1) was not present in this case, but the tidying up
1845 of the other signals was missing. I have made the two cases consistent.
1847 3. The ignore_target_hosts setting on a manualroute router was being ignored
1848 for hosts that were looked up using the /MX notation.
1850 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1853 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1854 operated on the sender address. After changing the $sender_address to <>
1855 for the sender address verify, Exim was re-instated it as the original
1856 (before rewriting) address, but remembering that it had rewritten it, so it
1857 wasn't rewriting it again. This bug also had the effect of breaking the
1858 sender address verification caching when the sender address was rewritten.
1860 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1861 This has been changed so that if the ip literal address matches
1862 ignore_target_hosts, the router declines.
1864 7. Added expansion conditions match_domain, match_address, and match_local_
1865 part (NOT match_host).
1867 8. The placeholder for the Received: header didn't have a length field set.
1869 9. Added code to Exim itself and to exim_lock to test for a specific race
1870 condition that could lead to file corruption when using MBX delivery. The
1871 issue is with the lockfile that is created in /tmp. If this file is removed
1872 after a process has opened it but before that process has acquired a lock,
1873 there is the potential for a second process to recreate the file and also
1874 acquire a lock. This could lead to two Exim processes writing to the file
1875 at the same time. The added code performs the same test as UW imapd; it
1876 checks after acquiring the lock that its file descriptor still refers to
1877 the same named file.
1879 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1880 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1883 11. Added the smtp_active_hostname option. If used, this will typically be made
1884 to depend on the incoming interface address. Because $interface_address is
1885 not set up until the daemon has forked a reception process, error responses
1886 that can happen earlier (such as "too many connections") no longer contain
1889 12. If an expansion in a condition on a "warn" statement fails because a lookup
1890 defers, the "warn" statement is abandoned, and the next ACL statement is
1891 processed. Previously this caused the whole ACL to be aborted.
1893 13. Added the iplsearch lookup type.
1895 14. Added ident_timeout as a log selector.
1897 15. Added tls_certificate_verified as a log selector.
1899 16. Added a global option tls_require_ciphers (compare the smtp transport
1900 option of the same name). This controls incoming TLS connections.
1902 17. I finally figured out how to make tls_require_ciphers do a similar thing
1903 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1904 before starting the TLS session.
1906 18. Tabs are now shown as \t in -bP output.
1908 19. If the log selector return_path_on_delivery was set, Exim crashed when
1909 bouncing a message because it had too many Received: header lines.
1911 20. If two routers both had headers_remove settings, and the first one included
1912 a superfluous trailing colon, the final name in the first list and the
1913 first name in the second list were incorrectly joined into one item (with a
1914 colon in the middle).
1920 1. Added -C and -D options to the exinext utility, mainly to make it easier
1921 to include in the automated testing, but these could be helpful when
1922 multiple configurations are in use.
1924 2. The exinext utility was not formatting the output nicely when there was
1925 an alternate port involved in the retry record key, nor when there was a
1926 message id as well (for retries that were specific to a specific message
1927 and a specific host). It was also confused by IPv6 addresses, because of
1928 the additional colons they contain. I have fixed the IPv4 problem, and
1929 patched it up to do a reasonable job for IPv6.
1931 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1932 delivery, the log line now contains "pipelined" if PIPELINING was used.
1934 4. An SMTP transport process used to panic and die if the bind() call to set
1935 an explicit outgoing interface failed. This has been changed; it is now
1936 treated in the same way as a connect() failure.
1938 5. A reference to $sender_host_name in the part of a conditional expansion
1939 that was being skipped was still causing a DNS lookup. This no longer
1942 6. The def: expansion condition was not recognizing references to header lines
1943 that used bh_ and bheader_.
1945 7. Added the _cache feature to named lists.
1947 8. The code for checking quota_filecount in the appendfile transport was
1948 allowing one more file than it should have been.
1950 9. For compatibility with Sendmail, the command line option
1958 and sets the incoming protocol and host name (for trusted callers). The
1959 host name and its colon can be omitted when only the protocol is to be set.
1960 Note the Exim already has two private options, -pd and -ps, that refer to
1961 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1962 "s", but I don't think that's a major issue.
1964 10. A number of refactoring changes to the code, none of which should affect
1967 (a) The number of logging options was getting close to filling up the
1968 32-bit word that was used as a bit map. I have split them into two classes:
1969 those that are passed in the argument to log_write(), and those that are
1970 only ever tested independently outside of that function. These are now in
1971 separate 32-bit words, so there is plenty of room for expansion again.
1972 There is no change in the user interface or the logging behaviour.
1974 (b) When building, for example, log lines, the code previously used a
1975 macro that called string_cat() twice, in order to add two strings. This is
1976 not really sufficiently general. Furthermore, there was one instance where
1977 it was actually wrong because one of the argument was used twice, and in
1978 one call a function was used. (As it happened, calling the function twice
1979 did not affect the overall behaviour.) The macro has been replaced by a
1980 function that can join an arbitrary number of extra strings onto a growing
1983 (c) The code for expansion conditions now uses a table and a binary chop
1984 instead of a serial search (which was left over from when there were very
1985 few conditions). Also, it now recognizes conditions like "pam" even when
1986 the relevant support is not compiled in: a suitably worded error message is
1987 given if an attempt is made to use such a condition.
1989 11. Added ${time_interval:xxxxx}.
1991 12. A bug was causing one of the ddress fields not to be passed back correctly
1992 from remote delivery subprocesses. The field in question was not being
1993 subsequently used, so this caused to problems in practice.
1995 13. Added new log selectors queue_time and deliver_time.
1997 14. Might have fixed a bug in maildirsizefile handling that threw up
1998 "unexpected character" debug warnings, and recalculated the data
1999 unnecessarily. In any case, I expanded the warning message to give more
2002 15. Added the message "Restricted characters in address" to the statements in
2003 the default ACL that block characters like @ and % in local parts.
2005 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2006 Three changes have been made:
2008 (a) There was a serious bug; a negative response to MAIL caused the whole
2009 recipient domain to be cached as invalid, thereby blocking all messages
2010 to all local parts at the same domain, from all senders. This bug has
2011 been fixed. The domain is no longer cached after a negative response to
2012 MAIL if the sender used is not empty.
2014 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2017 (c) A new callout option, "use_sender" has been added for people who want
2018 the modified behaviour.
2024 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2027 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2028 indeed breaks things for older releases.
2030 3. Added additional logging to the case where there is a problem reading data
2031 from a filter that is running in a subprocess using a pipe, in order to
2032 try to track down a specific problem.
2034 4. Testing facility fudge: when running in the test harness and attempting
2035 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2036 getting "No route to host". Convert this to a timeout.
2038 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2041 6. Some OS don't have socklen_t but use size_t instead. This affects the
2042 fifth argument of getsockopt() amongst other things. This is now
2043 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2044 can be set for individual OS. I have set it for SunOS5, OSF1, and
2045 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2046 some earlier ones do not.
2048 7. Change 4.30/15 was not doing the test caselessly.
2050 8. The standard form for an IPv6 address literal was being rejected by address
2051 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2052 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2053 this, as well as the form without the "IPv6" on the front (but only when
2054 address literals are enabled, of course).
2056 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2058 10. Exim crashed if a message with an empty sender address specified by -f
2059 encountered a router with an errors_to setting. This could be provoked only
2060 by a command such as
2064 where an empty string was supplied; "<>" did not hit this bug.
2066 11. Installed PCRE release 4.5.
2068 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2069 remained set. It is now erased.
2071 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2072 times from message ids (which are base 36 rather than the normal 62).
2074 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2075 were being counted as actual protocol errors, and logged if the log
2076 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2077 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2078 and DATA following a set of rejected RCPTs do not count as protocol errors.
2079 In other words, Exim assumes they were pipelined, though this may not
2080 actually be the case. Of course, in all cases the client gets an
2081 appropriate error code.
2083 15. If a lookup fails in an ACL condition, a message about the failure may
2084 be available; it is used if testing the ACL cannot continue, because most
2085 such messages specify what the cause of the deferral is. However, some
2086 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2087 that caused an old message to be retained and used if a later statement
2088 caused a defer, replacing the real cause of the deferral.
2090 16. If an IP address had so many PTR records that the DNS lookup buffer
2091 was not large enough to hold them, Exim could crash while trying to process
2092 the truncated data. It now detects and logs this case.
2094 17. Further to 4.21/58, another change has been made: if (and only if) the
2095 first line of a message (the first header line) ends with CRLF, a bare LF
2096 in a subsequent header line has a space inserted after it, so as not to
2097 terminate the header.
2099 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2100 unnecessarily, used atoi() instead of strtol(), and didn't check the
2101 termination when getting file sizes from file names by regex.
2103 19. Completely re-implemented the support for maildirsize files, in the light
2104 of a number of problems with the previous contributed implementation
2105 (4.30/29). In particular:
2107 . If the quota is zero, the maildirsize file is maintained, but no quota is
2110 . If the maildir directory does not exist, it is created before any attempt
2111 to write a maildirsize file.
2113 . The quota value in the file is just a cache; if the quota is changed in
2114 the transport, the new value overrides.
2116 . A regular expression is available for excluding directories from the
2119 20. The autoreply transport checks the characters in options that define the
2120 message's headers; it allows continued headers, but it was checking with
2121 isspace() after an embedded newline instead of explicitly looking for a
2124 21. If all the "regular" hosts to which an address was routed had passed their
2125 expiry times, and had not reached their retry times, the address was
2126 bounced, even if fallback hosts were defined. Now Exim should go on to try
2129 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2130 equivalent code in the SMTP transport. Some hosts send humungous responses
2131 to HELO/EHLO, more than 1024 it seems.
2133 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2134 gives compiler warnings in some environments. I've now done it "properly",
2137 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2138 (because of problems with the built-in one) was declared to return uschar *
2139 instead of char *, causing compiler failure.
2141 25. Fixed a file descriptor leak when processing alias/forward files.
2143 26. Fixed a minor format string issue in dbfn.c.
2145 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2147 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2148 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2151 29. When a sender address is verified, it is cached, to save repeating the test
2152 when there is more than one recipient in a message. However, when the
2153 verification involves a callout, it is possible for different callout
2154 options to be set for different recipients. It is too complicated to keep
2155 track of this in the cache, so now Exim always runs a verification when a
2156 callout is required, relying on the callout cache for the optimization.
2157 The overhead is duplication of the address routing, but this should not be
2160 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2161 to be verified with callout=postmaster, and the main callout worked but the
2162 postmaster check failed, the verification correctly failed. However, if a
2163 subsequent RCPT command asked for sender verification *without* the
2164 postmaster check, incorrect caching caused this verification also to fail,
2167 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2168 it was not caching the DNS options (qualify_single, search_parents) that
2169 were used when the lookup failed. A subsequent lookup with different
2170 options therefore always gave the same answer, though there were cases
2171 where it should not have. (Example: a "domains = !$mx_any" option on a
2172 dnslookup router: the "domains" option is always processed without any
2173 widening, but the router might have qualify_single set.) Now Exim uses the
2174 cached value only when the same options are set.
2176 32. Added John Jetmore's "exipick" utility to the distribution.
2178 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2179 than a timeout (e.g. a certificate is required, and is not provided), an
2180 Exim server now closes the connection immediately. Previously it waited for
2181 the client to close - but if the client is SSL, it seems that they each
2182 wait for each other, leading to a delay before one of them times out.
2184 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2185 maintained 0.8.x compatibility because I don't think many are using it, and
2186 it is clearly obsolete.
2188 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2191 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2192 client certificate was expired. A simple patch fixes this, though I don't
2193 understand the full logic of why the verify callback is called multiple
2196 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2197 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2198 which causes problems with some clients (such as the Certicom SSL Plus
2199 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2200 disables the coutermeasure allowing Eudora to connect."
2202 38. Exim was not checking that a write() to a log file succeeded. This could
2203 lead to Bad Things if a log got too big, in particular if it hit a file
2204 size limit. Exim now panics and dies if it cannot write to a log file, just
2205 as it does if it cannot open a log file.
2207 39. Modified OS/Makefile-Linux so that it now contains
2209 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2211 The two -D definitions ensure that Exim is compiled with large file
2212 support, which makes it possible to handle log files that are bigger than
2215 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2216 instance) a domain was checked against a named list that involved a lookup,
2217 causing $domain_data to be set, then another domain was checked against the
2218 same list, then the first domain was re-checked, the value of $domain_data
2219 after the final check could be wrong. In particular, if the second check
2220 failed, it could be set empty. This bug probably also applied to
2223 41. The strip_trailing_dot option was not being applied to the address given
2224 with the -f command-line option.
2226 42. The code for reading a message's header from the spool was incrementing
2227 $received_count, but never initializing it. This meant that the value was
2228 incorrect (doubled) while delivering a message in the same process in which
2229 it was received. In the most common configuration of Exim, this never
2230 happens - a fresh exec is done - but it can happen when
2231 deliver_drop_privilege is set.
2233 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2234 it now includes up to 150 characters of the unexpected data in the log
2237 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2238 and building data strings. The size of both of these buffers was 10 000
2239 bytes - far larger than anybody would *ever* want, thought I. Needless to
2240 say, somebody hit the limit. I have increased the maximum line length to
2241 20 000 and the maximum data length of concatenated lines to 100 000. I have
2242 also fixed two bugs, because there was no checking on these buffers. Tsk,
2243 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2244 buffer is too small.
2246 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2247 lsearch lookups. Now it does.
2249 46. When parsing a route_list item in a manualroute router, a fixed-length
2250 buffer was used for the list of hosts. I made this 1024 bytes long,
2251 thinking that nobody would ever have a list of hosts that long. Wrong.
2252 Somebody had a whole pile of complicated expansion conditions, and the
2253 string was silently truncated, leading to an expansion error. It turns out
2254 that it is easier to change to an unlimited length (owing to other changes
2255 that have happened since this code was originally written) than to build
2256 structure for giving a limitation error. The length of the item that
2257 expands into the list of hosts is now unlimited.
2259 47. The lsearch lookup could not handle data where the length of text line was
2260 more than 4095 characters. Such lines were truncated, leading to shortened
2261 data being returned. It should now handle lines of any length.
2263 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2264 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2267 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2268 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2269 no longer happens in comment lines. A list of replacements is now placed
2270 at the head of all of the source files, except those whose only change is
2271 to replace PERL_COMMAND in the very first #! line.
2273 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2274 messages on the queue, with a bottom-up merge sort, using code contributed
2275 by Michael Haardt. This should make operations like -bp somewhat faster on
2276 large queues. It won't affect queue runners, except when queue_run_in_order
2279 51. Installed eximstats 1.31 in the distribution.
2281 52. Added support for SRV lookups to the dnslookup router.
2283 53. If an ACL referred to $message_body or $message_body_end, the value was not
2284 reset for any messages that followed in the same SMTP session.
2286 54. The store-handling optimization for building very long strings was not
2287 differentiating between the different store pools. I don't think this
2288 actually made any difference in practice, but I've tidied it.
2290 55. While running the routers to verify a sender address, $sender_address
2291 was still set to the sender address. This is wrong, because when routing to
2292 send a bounce to the sender, it would be empty. Therefore, I have changed
2293 it so that, while verifying a sender address, $sender_address is set to <>.
2294 (There is no change to what happens when verifying a recipient address.)
2296 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2297 target A or AAAA records (if not already returned) without resetting the
2298 qualify_single or search_parents options of the DNS resolver. These are
2299 inappropriate in this case because the targets of MX and SRV records must
2300 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2301 target that, when qualified, matched something in the local domain. These
2302 two options are now turned off when doing these lookups.
2304 57. It seems that at least some releases of Reiserfs (which does not have the
2305 concept of a fixed number of inodes) returns zero and not -1 for the
2306 number of available inodes. This interacted badly with check_spool_inodes,
2307 which assumed that -1 was the "no such thing" setting. What I have done is
2308 to check that the total number of inodes is greater than zero before doing
2309 the test of how many are available.
2311 58. When a "warn" ACL statement has a log_message modifier, the message is
2312 remembered, and not repeated. This is to avoid a lot of repetition when a
2313 message has many recipients that cause the same warning to be written.
2314 Howewer, Exim was preserving the list of already written lines for an
2315 entire SMTP session, which doesn't seem right. The memory is now reset if a
2316 new message is started.
2318 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2319 debugging output unless log_rewrite was also set.
2321 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2322 of (int)(handle) when we know that handle contains (void *)(-1).
2324 61. The Exim daemon panic-logs an error return when it closes the incoming
2325 connection. However "connection reset by peer" seems to be common, and
2326 isn't really an error worthy of noting specially, so that particular error
2329 62. When Exim is trying to find all the local interfaces, it used to panic and
2330 die if the ioctl to get the interface flags failed. However, it seems that
2331 on at least one OS (Solaris 9) it is possible to have an interface that is
2332 included in the list of interfaces, but for which you get a failure error
2333 for this call. This happens when the interface is not "plumbed" into a
2334 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2335 failure of the "get flags" call assumes that the interface is down.
2337 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2338 makes life easier for people who are doing arithmetic on fields extracted
2339 from dates, where you often get leading zeros that should not be
2340 interpreted as octal.
2342 64. Added qualify_domain to the redirect router, to override the global
2345 65. If a pathologically long header line contained very many addresses (the
2346 report of this problem mentioned 10 000) and each of them was rewritten,
2347 Exim could use up a very large amount of memory. (It kept on making new
2348 copies of the header line as it rewrote, and never released the old ones.)
2349 At the expense of a bit more processing, the header rewriting function has
2350 been changed so that it no longer eats memory in this way.
2352 66. The generation of the Received: header has been moved from the time that a
2353 message starts to be received, to the time that it finishes. The timestamp
2354 in the Received: header should now be very close to that of the <= log
2355 line. There are two side-effects of this change:
2357 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2358 logged header lines no longer include the local Received: line, because
2359 it has not yet been created. The same applies to a copy of the message
2360 that is returned to a non-SMTP sender when a message is rejected.
2362 (b) When a filter file is tested using -bf, no additional Received: header
2363 is added to the test message. After some thought, I decided that this
2366 This change does not affect the value of $received_for. It is still set
2367 after address rewriting, but before local_scan() is called.
2369 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2371 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2372 gave an unhelpful panic error message, and a defer error. I have managed to
2373 change this behaviour so that it now rejects any supplied certificate,
2374 which seems right, as the list of acceptable certificates is empty.
2376 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2377 gave an unhelpful defer error. I have not managed to make this reject any
2378 supplied certificates, but the error message it gives is "no certificate
2379 supplied", which is not helpful.
2381 70. exigrep's output now also includes lines that are not associated with any
2382 message, but which match the given pattern. Implemented by a patch from
2383 Martin Sluka, which also tidied up the Perl a bit.
2385 71. Recipient callout verification, like sender verification, was using <> in
2386 the MAIL FROM command. This isn't really the right thing, since the actual
2387 sender may affect whether the remote host accepts the recipient or not. I
2388 have changed it to use the actual sender in the callout; this means that
2389 the cache record is now keyed on a recipient/sender pair, not just the
2390 recipient address. There doesn't seem to be a real danger of callout loops,
2391 since a callout by the remote host to check the sender would use <>.
2392 [SEE ABOVE: changed after hitting problems.]
2394 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2395 temporary errors. However, in the case of such a code being given after
2396 the end of a data transmission (i.e. after ".") Exim was failing to write
2397 a retry record for the message. (Yes, there was some broken host that was
2398 actually sending 8xx at this point.)
2400 73. An unknown lookup type in a host list could cause Exim to panic-die when
2401 the list was checked. (An example that provoked this was putting <; in the
2402 middle of a list instead of at the start.) If this happened during a DATA
2403 ACL check, a -D file could be left lying around. This kind of configuration
2404 error no longer causes Exim to die; instead it causes a defer errror. The
2405 incident is still logged to the main and panic logs.
2407 74. Buglet left over from Exim 3 conversion. The message "too many messages
2408 in one connection" was written to the rejectlog but not the mainlog, except
2409 when address rewriting (yes!) was being logged.
2411 75. Added write_rejectlog option.
2413 76. When a system filter was run not as root (that is, when system_filter_user
2414 was set), the values of the $n variables were not being returned to the
2415 main process; thus, they were not subsequently available in the $sn
2418 77. Added +return_path_on_delivery log selector.
2420 78. A connection timeout was being treated differently from recipients deferred
2421 when testing hosts_max_try with a message that was older than the host's
2422 retry timeout. (The host should not be counted, thus allowing all hosts to
2423 be tried at least once before bouncing.) This may have been the cause of an
2424 occasionally reported bug whereby a message would remain on the queue
2425 longer than the retry timeout, but would be bounced if a delivery was
2426 forced. I say "may" because I never totally pinned down the problem;
2427 setting up timeout/retry tests is difficult. See also the next item.
2429 79. The ultimate address timeout was not being applied to errors that involved
2430 a combination of host plus message (for example, a timeout on a MAIL
2431 command). When an address resolved to a number of possible hosts, and they
2432 were not all tried for each delivery (e.g. because of hosts_max_try), a
2433 message could remain on the queue longer than the retry timeout.
2435 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2438 81. Fixed an obscure SMTP outgoing bug which required at least the following
2439 conditions: (a) there was another message waiting for the same server;
2440 (b) the server returned 5xx to all RCPT commands in the first message so
2441 that the message was not completed; (c) the server dropped the connection
2442 or gave a negative response to the RSET that Exim sends to abort the
2443 transaction. The observed case was a dropped connection after DATA that had
2444 been sent in pipelining mode. That is, the server had advertised PIPELINING
2445 but was not implementing it correctly. The effect of the bug was incorrect
2446 behaviour, such as trying another host, and this could lead to a crash.
2452 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2453 and daemon.c were passed as pointers to ints; they should have been
2454 pointers to socklen_t variables (which are typically unsigned ints).
2456 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2459 3. Fixed a really odd bug that affected only the testing scheme; patching a
2460 certain fixed string in the binary changed the value of another string that
2461 happened to be identical to the end of the original first string.
2463 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2464 name", it returns that address as the IP address. On some operating
2465 systems (e.g. Solaris), it also passes back the IP address string as the
2466 "host name". However, on others (e.g. Linux), it passes back an empty
2467 string. Exim wasn't checking for this, and was changing the host name to an
2468 empty string, assuming it had been canonicized.
2470 5. Although rare, it is permitted to have more than one PTR record for a given
2471 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2472 all the names associated with an address, because they do in Solaris.
2473 However, it seems that they do not in Linux for data that comes from the
2474 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2475 I found this out when I moved to a new Linux workstation and tried to run
2476 the Exim test suite.
2478 To get round this problem I have changed the code so that it now does its
2479 own call to the DNS to look up PTR records when searching for a host name.
2480 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2481 addresses that are only in /etc/hosts are still found.
2483 This behaviour is, however, controlled by an option called host_lookup_
2484 order, which defaults to "bydns:byaddr". If people want to use the other
2485 order, or indeed, just use one or the other means of lookup, they can
2486 specify it in this variable.
2488 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2489 some operating systems, this comes back from gethostbyaddr() as an empty
2490 string, and this is what Exim used to test for. However, it seems that in
2491 other systems, "." is yielded. Exim now tests for this case too.
2493 7. The values of check_spool_space and check_log_space are now held internally
2494 as a number of kilobytes instead of an absolute number of bytes. If a
2495 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2496 kilobyte. This means that much larger values can be stored.
2498 8. Exim monitor: an attempt to get the action menu when not actually pointing
2499 at a message produces an empty menu entitled "No message selected". This
2500 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2501 no entries in it ("Shell widget menu has zero width and/or height"). So I
2502 have added a single, blank menu entry in this case.
2504 9. Added ${quote_local_part.
2506 10. MIME decoding is now applied to the contents of Subject: header lines when
2509 11. Now that a reference to $sender_host_address automatically causes a reverse
2510 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2511 host lookup before query-style lookups in lists that might use this
2512 variable. This has therefore been abolished, and the "net-" prefix is no
2513 longer necessary for query-style lookups.
2515 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2516 have been a typo for LFLAGS, so it has been changed.
2518 13. The install script calls Exim with "-C /dev/null" in order to find the
2519 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2520 to be output. Howeve, since Exim outputs its version number before the
2521 error, it didn't break the script. It just looked ugly. I fixed this by
2522 always allowing "-C /dev/null" if the caller is root.
2524 14. Ignore overlarge ACL variable number when reading spool file - insurance
2525 against a later release with more variables having written the file.
2527 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2528 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2529 this, as well as the form without the "IPv6" on the front.
2531 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2532 OS/Makefile-Darwin file.
2534 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2535 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2537 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2538 variable. However, some OS use ints for inodes. I've added cast to long int
2539 to get rid of the compiler warning.
2541 19. I had forgotten to lock out "/../" in configuration file names when
2542 ALT_CONFIG_PREFIX was set.
2544 20. Routers used for verification do not need to specify transports. However,
2545 if such a router generated a host list, and callout was configured, Exim
2546 crashed, because it could not find a port number from the (non-existent)
2547 transport. It now assumes port 25 in this circumstance.
2549 21. Added the -t option to exigrep.
2551 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2552 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2553 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2554 them is set, LOOKUP_LSEARCH is forced.
2556 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2558 24. Added sender and host information to the "rejected by local_scan()" log
2559 line; previously there was no indication of these.
2561 25. Added .include_if_exists.
2563 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2564 Linux. It turns out that not all file systems support this. Apparently some
2565 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2566 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2567 not supported on the file descriptor, is now ignored when Exim is trying to
2568 sync a directory. This applies only to Linux.
2570 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2572 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2573 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2576 29. Added support for maildirsize files from supplied patch (modified a bit).
2578 30. The use of :fail: followed by an empty string could lead Exim to respond to
2579 sender verification failures with (e.g.):
2581 550 Verification failed for <xxx>
2582 550 Sender verify failed
2584 where the first response line was missing the '-' that indicates it is not
2585 the final line of the response.
2587 31. The loop for finding the name of the user that called Exim had a hardwired
2588 limit of 10; it now uses the value of finduser_retries, which is used for
2589 all other user lookups.
2591 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2594 33. Exim was neglecting to zero errno before one call of strtol() when
2595 expanding a string and expecting an integer value. On some systems this
2596 resulted in spurious "integer overflow" errors. Also, it was casting the
2597 result into an int without checking.
2599 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2600 did not work. The code looks as if it has *never* worked, though it appears
2601 to have been documented since at least releast 1.62. I have made it work.
2603 35. The "timeout_DNS" error in retry rules, also documented since at least
2604 1.62, also never worked. As it isn't clear exactly what this means, and
2605 clearly it isn't a major issue, I have abolished the feature by treating it
2606 as "timeout", and writing a warning to the main and panic logs.
2608 36. The display of retry rules for -brt wasn't always showing the error code
2611 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2612 timeout_connect_A, timeout_connect_MX.
2614 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2615 to the empty sender.
2617 39. The daemon was not analysing the content of -oX till after it had closed
2618 stderr and disconnected from the controlling terminal. This meant that any
2619 syntax errors were only noted on the panic log, and the return code from
2620 the command was 0. By re-arranging the code a little, I've made the
2621 decoding happen first, so such errors now appear on stderr, and the return
2622 code is 1. However, the actual setting up of the sockets still happens in
2623 the disconnected process, so errors there are still only recorded on the
2626 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2627 connections (as happens on some IP stacks) was logged at start up time as
2628 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2629 it from "IPv6 and IPv4", which means that two separate sockets are being
2632 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2633 says whether AF_INET or AF_INET6 was passed as an argument.
2635 42. Exiwhat output was messed up when time zones were included in log
2638 43. Exiwhat now gives more information about the daemon's listening ports,
2639 and whether -tls-on-connect was used.
2641 44. The "port" option of the smtp transport is now expanded.
2643 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2644 silently ignored. Now an error message is written to the main and panic
2647 46. There's a new ACL modifier called "logwrite" which writes to a log file
2648 as soon as it is encountered.
2650 47. Added $local_user_uid and $local_user_gid at routing time.
2652 48. Exim crashed when trying to verify a sender address that was being
2655 49. Exim was recognizing only a space character after ".include". It now also
2656 recognizes a tab character.
2658 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2659 extracting the relevant information from the specification. The man page no
2660 longer contains scrambled data for the -d option, and I've added a section
2661 at the front about calling Exim under different names.
2663 51. Added "extra_headers" argument to the "mail" command in filter files.
2665 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2668 53. Installed eximstats 1.29.
2670 54. Added transport_filter_timeout as a generic transport option.
2672 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2673 Cc: header lines. This was required by RFC 822, but it not required by RFC
2676 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2677 incoming messages that did not have them. Now it does so only if the
2678 message originates locally, that is, if there is no associated remote host
2679 address. When Resent- header lines are present, this applies to the Resent-
2680 lines rather than the non-Resent- lines.
2682 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2683 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2685 58. Messages for configuration errors now include the name of the main
2686 configuration file - useful now that there may be more than one file in a
2687 list (.included file names were always shown).
2689 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2690 for those rare installations that do not start the daemon as root or run it
2691 setuid root. I've cut out the call to initgroups() if the daemon is not
2694 60. The Exim user and group can now be bound into the binary as text strings
2695 that are looked up at the start of Exim's processing.
2697 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2699 62. Added $mailstore_basename variable.
2701 63. Installed patch to sieve.c from Michael Haardt.
2703 64. When Exim failed to open the panic log after failing to open the main log,
2704 the original message it was trying to log was written to stderr and debug
2705 output, but if they were not available (the usual case in production), it
2706 was lost. Now it is written to syslog before the two lines that record the
2707 failures to open the logs.
2709 65. Users' Exim filters run in subprocesses under the user's uid. It is
2710 possible for a "deliver" command or an alias in a "personal" command to
2711 provoke an address rewrite. If logging of address rewriting is configured,
2712 this fails because the process is not running as root or exim. There may be
2713 a better way of dealing with this, but for the moment (because 4.30 needs
2714 to be released), I have disabled address rewrite logging when running a
2715 filter in a non-root, non-exim process.
2721 1. The buildconfig auxiliary program wasn't quoting the value set for
2722 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2723 not defined. This bug was masked in 4.22 by the effect that was fixed in
2726 2. Some messages that were rejected after a message id was allocated were
2727 shown as "incomplete" by exigrep. It no longer does this for messages that
2728 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2730 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2731 have allow_domain_literals set, the ID did not get logged in the <= line.
2732 Domain literals are now always recognized in Message-ID: header lines.
2734 4. The first argument for a ${extract expansion item is the key name or field
2735 number. Leading and trailing spaces in this item were not being ignored,
2736 causing some misleading effects.
2738 5. When deliver_drop_privilege was set, single queue runner processes started
2739 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2740 same command in the process it spins off) were not dropping privilege.
2742 6. When the daemon running as "exim" started a queue runner, it always
2743 re-executed Exim in the spun-off process. This is a waste of effort when
2744 deliver_drop_privilege is set. The new process now just calls the
2745 queue-runner function directly.
2751 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2754 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2755 ignored. Though the use of -oP was forcing the writing of a pid file, it
2756 was always written to the default place.
2758 3. If the message "no IP address found for host xxxx" is generated during
2759 incoming verification, it is now followed by identification of the incoming
2760 connection (so you can more easily find what provoked it).
2762 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2764 5. Added some features to "harden" Exim a bit more against certain attacks:
2766 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2767 be put in Local/Makefile. This is like the never_users runtime option,
2768 but it cannot be overridden. The default setting is "root".
2770 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2771 prefix string with which any file named in a -C command line option
2774 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2775 is retained for -C and -D only if the caller of Exim is root. Without
2776 it, the exim user may also use -C and -D and retain privilege.
2778 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2779 command line option is disabled.
2781 6. Macro names set by the -D option must start with an upper case letter, just
2782 like macro names defined in the configuration file.
2784 7. Added "dereference=" facility to LDAP.
2786 8. Two instances of the typo "uknown" in the source files are fixed.
2788 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2789 the Configure-Makefile script screwed up while processing it.
2791 10. Incorporated PCRE 4.4.
2793 11. The SMTP synchronization check was not operating right at the start of an
2794 SMTP session. For example, it could not catch a HELO sent before the client
2795 waited for the greeting. There is now a check for outstanding input at the
2796 point when the greeting is written. Because of the duplex, asynchronous
2797 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2798 way, but not yet received, when the check is performed.
2800 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2801 on TCP/IP sockets, because this apparently causes some broken clients to
2804 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2805 unchanged) from the Cygwin maintainer.
2807 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2808 was supported instead of testing for mbx. Effectively a typo.
2810 15. The spa authenticator server code was not checking that the input it
2811 received was valid base64.
2813 16. The debug output line for the "set" modifier in ACLs was not showing the
2814 name of the variable that was being set.
2816 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2818 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2819 lookup on its own. Something else was needed to trigger the lookup. For
2820 example, a match in host_lookup or the need for a host name in a host list.
2821 Now, if $sender_host_name is referenced and the host name has not yet been
2822 looked up, a lookup is performed. If the lookup fails, the variable remains
2823 empty, and $host_lookup_failed is set to "1".
2825 19. Added "eqi" as a case-independent comparison operator.
2827 20. The saslauthd authentication condition could segfault if neither service
2828 nor realm was specified.
2830 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2831 error message that was logged was misleading, and incoming SMTP
2832 connections were dropped. The message is now more accurate, and temporary
2833 errors are given to SMTP connections.
2835 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2836 (except RSET) with a 421 error, until QUIT is received. However, it was
2837 failing to send a response to QUIT.
2839 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2840 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2841 ACL is now run after the helo_try_verify_hosts code.
2843 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2844 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2845 case-independent, so other case variants are also recognized). Apparently
2846 some systems use these upper case variants.
2848 25. If more than two messages were waiting for the same host, and a transport
2849 filter was specified for the transport, Exim sent two messages over the
2850 same TCP/IP connection, and then failed with "socket operation on non-
2851 socket" when it tried to send the third.
2853 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2855 27. The extern definition of crypt16() in expand.c was not being excluded when
2856 the OS had its own crypt16() function.
2858 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2859 as a preferred synonym for return_size_limit, both as an option and as an
2862 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2864 30. Changed the default configuration ACL to relax the local part checking rule
2865 for addresses that are not in any local domains. For these addresses,
2866 slashes and pipe symbols are allowed within local parts, but the sequence
2867 /../ is explicitly forbidden.
2869 31. SPA server authentication was not clearing the challenge buffer before
2872 32. log_message in a "warn" ACL statement was writing to the reject log as
2873 well as to the main log, which contradicts the documentation and doesn't
2874 seem right (because no rejection is happening). So I have stopped it.
2876 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2877 However, I am unable to do any testing of this.
2879 34. Fixed an infelicity in the appendfile transport. When checking directories
2880 for a mailbox, to see if any needed to be created, it was accidentally
2881 using path names with one or more superfluous leading slashes; tracing
2882 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2884 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2885 contents are added to the log line that is written for every discarded
2886 recipient. (Previously a log_message setting was ignored.)
2888 36. The ${quote: operator now quotes the string if it is empty.
2890 37. The install script runs exim in order to find its version number. If for
2891 some reason other than non-existence or emptiness, which it checks, it
2892 could not run './exim', it was installing it with an empty version number,
2893 i.e. as "exim-". This error state is now caught, and the installation is
2896 38. An argument was missing from the function that creates an error message
2897 when Exim fails to connect to the socket for saslauthd authentication.
2898 This could cause Exim to crash, or give a corrupted message.
2900 39. Added isip, isip4, and isip6 to ${if conditions.
2902 40. The ACL variables $acl_xx are now saved with the message, and can be
2903 accessed later in routers, transports, and filters.
2905 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2906 strings in the file are not string-expanded.
2908 42. If a MAIL command specified a SIZE value that was too large to fit into an
2909 int variable, the check against message_size_limit failed. Such values are
2910 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2911 day this will have to be increased, but I don't think I want to be around
2912 when emails are that large.
2919 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2920 iconv() is not standard in FreeBSD.
2922 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2923 IPv6 enabled. The observed symptom was a segmentation fault on return from
2924 the function os_common_find_running_interfaces() in src/os.c.
2926 3. In the check_special_case() function in daemon.c I had used "errno" as an
2927 argument name, which causes warnings on some systems. This was basically a
2928 typo, since it was named "eno" in the comments!
2930 4. The code that waits for the clock to tick (at a resolution of some fraction
2931 of a second) so as to ensure message-id uniqueness was always waiting for
2932 at least one whole tick, when it could have waited for less. [This is
2933 almost certainly not relevant at current processor speeds, where it is
2934 unlikely to ever wait at all. But we try to future-proof.]
2936 5. The function that sleeps for a time interval that includes fractions of a
2937 second contained a race. It did not block SIGALRM between setting the
2938 timer, and suspending (a couple of lines later). If the interval was short
2939 and the sigsuspend() was delayed until after it had expired, the suspension
2940 never ended. On busy systems this could lead to processes getting stuck for
2943 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2944 process, before it forks any delivery processes. The open lookup caching
2945 mechanism meant that the open file or database connection was passed into
2946 the delivery process. The problem was that delivery processes always tidy
2947 up cached lookup data. This could cause a problem for the next delivery
2948 process started by the queue runner, because the external queue runner
2949 process does not know about the closure. So the next delivery process
2950 still has data in the lookup cache. In the case of a file lookup, there was
2951 no problem because closing a file descriptor in a subprocess doesn't affect
2952 the parent. However, if the lookup was caching a connection to a database,
2953 the connection was closed, and the second delivery process was likely to
2954 see errors such as "PGSQL: query failed: server closed the connection
2955 unexpectedly". The problem has been fixed by closing all cached lookups
2956 in a queue runner before running a delivery process.
2958 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2959 seem to have the "const" qualifier which it has on other OS. I've
2962 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2963 *of the same type* (client or server) with the same public name that an
2964 error should be diagnosed.
2966 9. When Exim looked up a host name for an IP address, but failed to find the
2967 original IP address when looking up the host name (a safety check), it
2968 output the message "<ip address> does not match any IP for NULL", which was
2969 confusing, to say the least. The bug was that the host name should have
2970 appeared instead of "NULL".
2972 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2973 user that is built into the binary, and the -C or -D options is used, root
2974 privilege is dropped before the configuration file is read. In addition,
2975 logging is switched to stderr instead of the normal log files. If the
2976 configuration then re-defines the Exim user, the unprivileged environment
2977 is probably not what is expected, so Exim logs a panic warning message (but
2980 However, if deliver_drop_privilege is set, the unprivileged state may well
2981 be exactly what is intended, so the warning has been cut out in that case,
2982 and Exim is allowed to try to write to its normal log files.
2988 1. smtp_return_error_details was not giving details for temporary sender
2989 or receiver verification errors.
2991 2. Diagnose a configuration error if two authenticators have the same public
2994 3. Exim used not to create the message log file for a message until the first
2995 delivery attempt. This could be confusing when incoming messages were held
2996 for policy or load reasons. The message log file is now created at the time
2997 the message is received, and an initial "Received" line is written to it.
2999 4. The automatically generated man page for command line options had a minor
3000 bug that caused no ill effects; however, a more serious problem was that
3001 the procedure for building the man page automatically didn't always
3002 operate. Consequently, release 4.20 contains an out-of-date version. This
3003 shouldn't happen again.
3005 5. When building Exim with embedded Perl support, the script that builds the
3006 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3007 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3009 6. The freeze_tell option was not being used for messages that were frozen on
3010 arrival, either by an ACL or by local_scan().
3012 7. Added the smtp_incomplete_transaction log selector.
3014 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3015 was accepting AUTH without a new EHLO.
3017 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3018 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3021 10. Logging of TCP/IP connections (when configured) now happens in the main
3022 daemon process instead of the child process, so that the TCP/IP connection
3023 count is more accurate (but it can never be perfect).
3025 11. The use of "drop" in a nested ACL was not being handled correctly in the
3026 outer ACL. Now, if condition failure induced by the nested "drop" causes
3027 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3028 or "require"), the connection is dropped.
3030 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3031 that yield "discard" can now be used with an "accept" or a "discard" verb,
3032 but an error is generated for any others (because I can't see a useful way
3033 to define what should happen).
3035 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3036 are now processed in the same way as lines in the Exim configuration file.
3037 In particular, continuation lines are supported.
3039 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3041 15. Added -ti meaning -t -i.
3043 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3044 domains, and warn by logging if others are found.
3046 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3047 returned from GIFCONF when Exim is trying to find the list of interfaces on
3048 a host. The code in os.c has been modified to copy each ifreq to an aligned
3049 structure in all cases.
3051 Also, in some cases, the returned ifreq's were being copied to a 'struct
3052 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3053 means the last couple of bytes of an IPv6 address could be chopped if the
3054 ifreq contained only a normal sockaddr (14 bytes storage).
3056 18. Named domain lists were not supported in the hosts_treat_as_local option.
3057 An entry such as +xxxx was not recognized, and was treated as a literal
3060 19. Ensure that header lines added by a DATA ACL are included in the reject log
3061 if the ACL subsequently rejects the message.
3063 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3064 MD5 (which is deprecated).
3066 21. When testing a filter file using -bf, Exim was writing a message when it
3067 took the sender from a "From " line in the message, but it was not doing so
3068 when it took $return_path from a Return-Path: header line. It now does.
3070 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3071 with a valid header line field name (a series of printing characters
3072 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3074 23. Changed "disc" in the source to "disk" to conform to the documentation and
3075 the book and for uniformity.
3077 24. Ignore Sendmail's -Ooption=value command line item.
3079 25. When execve() failed while trying to run a command in a pipe transport,
3080 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3081 could be confused with a return value of 69 from the command itself. This
3082 has been changed to 127, the value the shell returns if it is asked to run
3083 a non-existent command. The wording for the related log line suggests a
3084 non-existent command as the problem.
3086 26. If received_header_text expands to an empty string, do not add a Received:
3087 header line to the message. (Well, it adds a token one on the spool, but
3088 marks it "old" so that it doesn't get used or transmitted.)
3090 27. Installed eximstats 1.28 (addition of -nt option).
3092 28. There was no check for failure on the call to getsockname() in the daemon
3093 code. This can fail if there is a shortage of resources on the system, with
3094 ENOMEM, for example. A temporary error is now given on failure.
3096 29. Contrary to the C standard, it seems that in some environments, the
3097 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3098 program. Exim now does this explicitly; it affects the formatting of
3099 timestamps using strftime().
3101 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3102 complaints. I've now initialized all the variables, to avoid this.
3104 32. Header lines added by a system filter were not being "seen" during
3105 transport-time rewrites.
3107 33. The info_callback() function passed to OpenSSL is set up with type void
3108 (*)(SSL *, int, int), as described somewhere. However, when calling the
3109 function (actually a macro) that sets it up, the type void(*)() is
3110 expected. I've put in a cast to prevent warnings from picky compilers.
3112 34. If a DNS black list lookup found a CNAME record, but there were no A
3113 records associated with the domain it pointed at, Exim crashed.
3115 35. If a DNS black list lookup returned more than one A record, Exim ignored
3116 all but the first. It now scans all returned addresses if a particular IP
3117 value is being sought. In this situation, the contents of the
3118 $dnslist_value variable are a list of all the addresses, separated by a
3121 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3122 to accept a host name and all its aliases if the forward lookup for any of
3123 them yielded the IP address of the incoming connection. Now it accepts only
3124 those names whose forward lookup yields the correct IP address. Any other
3125 names are discarded. This closes a loophole whereby a rogue DNS
3126 administrator could create reverse DNS records to break through a
3127 wildcarded host restriction in an ACL.
3129 37. If a user filter or a system filter that ran in a subprocess used any of
3130 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3131 the wrong values were passed to the pipe command ($thisaddress had the
3132 value of $0, $0 had the value of $1, etc). This bug was introduced by
3133 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3135 38. Improved the line breaking for long SMTP error messages from ACLs.
3136 Previously, if there was no break point between 40 and 75 characters, Exim
3137 left the rest of the message alone. Two changes have been made: (a) I've
3138 reduced the minimum length to 35 characters; (b) if it can't find a break
3139 point between 35 and 75 characters, it looks ahead and uses the first one
3140 that it finds. This may give the occasional overlong line, but at least the
3141 remaining text gets split now.
3143 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3144 file descriptors might be low, and that setting 1000 would always raise it.
3145 It turns out that in some environments, the limit is already over 1000 and
3146 that lowering it causes trouble. So now Exim takes care not to decrease it.
3148 40. When delivering a message, the value of $return_path is set to $sender_
3149 address at the start of routing (routers may change the value). By an
3150 oversight, this default was not being set up when an address was tested by
3151 -bt or -bv, which affected the outcome if any router or filter referred to
3154 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3155 log only when "message" or "log_message" are set. However, if one of the
3156 conditions was an address verification, or a call to a nested ACL, the
3157 messages generated by the underlying test were being passed through. This
3158 no longer happens. The underlying message is available in $acl_verify_
3159 message for both "message" and "log_message" expansions, so it can be
3160 passed through if needed.
3162 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3163 new expansion $bh_ to give the encoded byte string without charset
3164 translation. Translation happens only if iconv() is available; HAVE_ICONV
3165 indicates this at build time. HEADERS_CHARSET gives the charset to
3166 translate to; headers_charset can change it in the configuration, and
3167 "headers charset" can change it in an individual filter file.
3169 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3170 that creates RFC 2047 encoded "words" labels them as that charset instead
3171 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3172 expansion operator; (ii) when Exim creates a From: line for a local
3173 message; (iii) when a header line is rewritten to include a "phrase" part.
3175 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3176 buggy, causing it to skip the first lines of messages whose message ID
3177 ended in 'D'. This would not have bitten before Exim release 4.14, because
3178 message IDs were unlikely to end in 'D' before then. The effect was to have
3179 incorrect size information for certain domains.
3181 45. #include "config.h" was missing at the start of the crypt16.c module. This
3182 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3185 46. If there was a timeout during a "random" callout check, Exim treated it as
3186 a failure of the random address, and carried on sending RSET and the real
3187 address. If the delay was just some slowness somewhere, the response to the
3188 original RCPT would be taken as a response to RSET and so on, causing
3189 mayhem of various kinds.
3191 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3192 when I implemented it. It didn't allow for the fact that some option values
3193 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3194 the right test for positive values.
3196 48. Domain names in DNS records are case-independent. Exim always looks them up
3197 in lower case. Some resolvers return domain names in exactly the case they
3198 appear in the zone file, that is, they may contain uppercase letters. Not
3199 all resolvers do this - some return always lower case. Exim was treating a
3200 change of case by a resolver as a change of domain, similar to a widening
3201 of a domain abbreviation. This triggered its re-routing code and so it was
3202 trying to route what was effectively the same domain again. This normally
3203 caused routing to fail (because the router wouldn't handle the domain
3204 twice). Now Exim checks for this case specially, and just changes the
3205 casing of the domain that it ultimately uses when it transmits the message
3208 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3211 50. If a filter generated a file delivery with a non-absolute name (possible if
3212 no home directory exists for the router), the forbid_file option was not
3215 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3216 the existing equality matching.
3218 52. Exim was using ints instead of ino_t variables in some places where it was
3219 dealing with inode numbers.
3221 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3222 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3223 called TMPDIR, and if it finds it is different, it changes its value.
3225 54. The smtp_printf() function is now made available to local_scan() so
3226 additional output lines can be written before returning. There is also an
3227 smtp_fflush() function to enable the detection of a dropped connection.
3228 The variables smtp_input and smtp_batched_input are exported to
3231 55. Changed the default runtime configuration: the message "Unknown user"
3232 has been removed from the ACL, and instead placed on the localuser router,
3233 using the cannot_route_message feature. This means that any verification
3234 failures that generate their own messages won't get overridden. Similarly,
3235 the "Unrouteable address" message that was in the ACL for unverifiable
3236 relay addresses has also been removed.
3238 56. Added hosts_avoid_esmtp to the smtp transport.
3240 57. The exicyclog script was not checking for the esoteric option
3241 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3242 will work only if exicyclog is run under the appropriate euid.
3244 58. Following a discussion on the list, the rules by which Exim recognises line
3245 endings on incoming messages have been changed. The -dropcr and drop_cr
3246 options are now no-ops, retained only for backwards compatibility. The
3247 following line terminators are recognized: LF CRLF CR. However, special
3248 processing applies to CR:
3250 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3251 nor a local message in the state where . is a terminator.
3253 (ii) If a bare CR is encountered in a header line, an extra space is added
3254 after the line terminator so as not to end the header. The reasoning
3255 behind this is that bare CRs in header lines are most likely either
3256 to be mistakes, or people trying to play silly games.
3258 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3259 was being incorrectly given as 18 bytes larger than it should have been.
3260 This is a VOB (very old bug).
3262 60. This may never have affected anything current, but just in case it has:
3263 When the local host is found other than at the start of a list of hosts,
3264 the local host, those with the same MX, and any that follow, are discarded.
3265 When the list in question was part of a longer list of hosts, the following
3266 hosts (not currently being processed) were also being discarded. This no
3267 longer happens. I'm not sure if this situation could ever has previously
3270 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3273 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3274 header. This is something that is recommended in a new Internet Draft, and
3275 is something that is documented as being done by Sendmail. There are two
3276 possible values. For messages generated by the autoreply transport, Exim
3279 Auto-Submitted: auto-replied
3281 whereas for all other generated messages (e.g. bounces) it adds
3283 Auto-Submitted: auto-generated
3285 63. The "personal" condition in filters now includes a test for the
3286 Auto-Submitted: header. If it contains the string "auto-" the message it
3287 not considered personal.
3289 64. Added rcpt_include_affixes as a generic transport option.
3291 65. Added queue_only_override (default true).
3293 66. Added the syslog_duplication option.
3295 67. If what should have been the first header line of a message consisted of
3296 a space followed by a colon, Exim was mis-interpreting it as a header line.
3297 It isn't of course - it is syntactically invalid and should therefore be
3298 treated as the start of the message body. The misbehaviour could have
3299 caused a number of strange effects, including loss of data in subsequent
3300 header lines, and spool format errors.
3302 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3303 client host had authenticated. This control can now be exercised by an ACL
3304 for more flexibility.
3306 69. By default, callouts do not happen when testing with -bh. There is now a
3307 variant, -bhc, which does actually run the callout code, including
3308 consulting and updating the callout cache.
3310 70. Added support for saslauthd authentication, courtesy of Alexander
3313 71. If statvfs() failed on the spool or log directories while checking their
3314 size for availability, Exim confusingly gave the error "space shortage".
3315 Furthermore, in debugging mode it crashed with a floating point exception.
3316 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3317 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3318 serious problem, Exim now writes to the main and panic logs when this
3319 happens, with details of the failure. It then refuses to accept the
3320 incoming message, giving the message "spool directory problem" or "log
3321 directory problem" with a 421 code for SMTP messages.
3323 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3324 0, 1, and 2 exist, because some OS complain for execs without them (see
3325 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3326 descriptors. However, the code omitted to check that the open succeeded,
3327 causing mysterious errors if for some reason the permissions on /dev/null
3328 got screwed. Now Exim writes a message to the main and panic logs, and
3329 bombs out if it can't open /dev/null.
3331 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3332 interact so that it is all more flexible. It is supposed to remain
3333 backwards compatible. Also added extra_local_interfaces.
3335 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3336 to bomb out with an assertion failure - to the client this appears as a
3337 connection drop. This problem occurs in the part of the code that was taken
3338 from the Samba project. Fortunately, the assertion is in a very simple
3339 function, so I have fixed this by reproducing the function inline in the
3340 one place where it is called, and arranging for authentication to fail
3341 instead of killing the process with assert().
3343 75. The SPA client code was not working when the server requested OEM rather
3344 than Unicode encoding.
3346 76. Added code to make require_files with a specific uid setting more usable in
3347 the case where statting the file as root fails - usually a non-root-mounted
3348 NFS file system. When this happens and the failure is EACCES, Exim now
3349 forks a subprocess and does the per-uid checking as the relevant uid.
3351 77. Added process_log_path.
3353 78. If log_file_path was not explicitly set, a setting of check_log_space or
3354 check_log_inodes was ignored.
3356 79. If a space check for the spool or log partitions fails, the incident is now
3357 logged. Of course, in the latter case the data may get lost...
3359 80. Added the %p formatting code to string_format() so that it can be used to
3360 print addresses in debug_print(). Adjusted all the address printing in the
3361 debugging in store.c to use %p rather than %d.
3363 81. There was a concern that a line of code in smtp_in.c could overflow a
3364 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3365 initially expressed, the concern was not well-founded, because trailing
3366 spaces are removed early. However, if the trailing spaces were followed by
3367 a NULL, they did not get removed, so the overflow was possible. Two fixes
3370 (a) I re-wrote the offending code in a cleaner fashion.
3371 (b) If an incoming SMTP command contains a NULL character, it is rejected
3374 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3375 runs initgroups(), so that if the Exim user is in any additional groups,
3376 they will be used during message reception.
3382 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob