1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.226 2005/09/13 10:35:52 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.
186 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
188 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
189 command itself ended in error, the underlying message about the transport
190 filter timeout was being overwritten with the pipe command error. Now the
191 underlying error message should be appended to the second error message.
193 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
195 PH/02 Reduce the amount of output that "make" produces by default. Full output
196 can still be requested.
198 PH/03 The warning log line about a condition test deferring for a "warn" verb
199 was being output only once per connection, rather than after each
200 occurrence (because it was using the same function as for successful
201 "warn" verbs). This seems wrong, so I have changed it.
203 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
204 it should not have, which might have caused a crash in the right
205 circumstances, but probably never did.
207 PH/04 Installed a modified version of Tony Finch's patch to make submission
208 mode fix the return path as well as the Sender: header line, and to
209 add a /name= option so that you can make the user's friendly name appear
212 TF/03 Added the control = fakedefer ACL modifier.
214 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
215 Mark Lowes for thorough testing.
217 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
219 TK/03 Merged latest SRS patch from Miles Wilton.
221 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
222 with the definition in sysexits.h (which is #included earlier).
223 Fortunately, Exim does not actually use EX_OK. The code used to try to
224 preserve the sysexits.h value, by assumimg that macro definitions were
225 scanned for macro replacements. I have been disabused of this notion,
226 so now the code just undefines EX_OK before #including unistd.h.
228 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
229 in the smtp transport. When a block could not be written in a single
230 write() function, the timeout was being re-applied to each part-write.
231 This seems wrong - if the receiver was accepting one byte at a time it
232 would take for ever. The timeout is now adjusted when this happens. It
233 doesn't have to be particularly precise.
235 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
236 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
238 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
239 without a domain if the check to <postmaster@domain> fails.
241 SC/01 Eximstats: added -xls and the ability to specify output files
242 (patch written by Frank Heydlauf).
244 SC/02 Eximstats: use FileHandles for outputing results.
246 SC/03 Eximstats: allow any combination of xls, txt, and html output.
248 SC/04 Eximstats: fixed display of large numbers with -nvr option
250 SC/05 Eximstats: fixed merging of reports with empty tables.
252 SC/06 Eximstats: added the -include_original_destination flag
254 SC/07 Eximstats: removed tabs and trailing whitespace.
256 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
258 TK/06 MBOX spool code: Add real "From " MBOX separator line
259 so the .eml file is really in mbox format (even though
260 most programs do not really care). Patch from Alex Miller.
262 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
263 The latter is generated from $received_to and is only set if the
264 message has one envelope recipient. SA can use these headers,
265 obviously out-of-the-box. Patch from Alex Miller.
267 PH/08 The ${def test on a variable was returning false if the variable's
268 value was "0", contrary to what the specification has always said!
269 The result should be true unless the variable is empty.
271 PH/09 The syntax error of a character other than { following "${if
272 def:variable_name" (after optional whitespace) was not being diagnosed.
273 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
274 accidental colon was present, for example, could give incorrect results.
276 PH/10 Tidied the code in a number of places where the st_size field of a stat()
277 result is used (not including appendfile, where other changes are about
280 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
281 This involved changing a lot of size variables from int to off_t. It
282 should work with maildirs and everything.
284 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
285 spamd dying while we are connected to it.
287 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
290 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
291 function for BDB 4.3.
293 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
295 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
296 the sources for PCRE have been re-arranged and more files are now
299 PH/15 The code I had for printing potentially long long variables in PH/11
300 above was not the best (it lost precision). The length of off_t variables
301 is now inspected at build time, and an appropriate printing format (%ld
302 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
303 to be "long long int" or "long int". This is needed for the internal
304 formatting function string_vformat().
306 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
307 the configuration file to be ":syslog", then the script "guesses" where
308 the logs files are, rather than using the compiled in default. In our
309 case the guess is not the same as the compiled default, so the script
310 suddenly stopped working when I started to use syslog. The patch checks
311 to see if log_file_path is "". If so, it attempts to read it from exim
312 with no configuration file to get the compiled in version, before it
313 falls back to the previous guessing code."
315 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
316 implementing BATV in an Exim configuration. See NewStuff for the gory
319 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
320 Makefile that are specific to HP-UX.
322 PH/18 If the "use_postmaster" option was set for a recipient callout together
323 with the "random" option, the postmaster address was used as the MAIL
324 FROM address for the random test, but not for the subsequent recipient
325 test. It is now used for both.
327 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
328 patch removes a few documentation additions to RFC 3028, because the
329 latest draft now contains them. It adds the new en;ascii-case comparator
330 and a new error check for 8bit text in MIME parts. Comparator and
331 require names are now matched exactly. I enabled the subaddress
332 extension, but it is not well tested yet (read: it works for me)."
334 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
335 rework some of the code of TK/09 above to avoid the hardwired use of
336 "%lld" and "long long". Replaced the call to snprintf() with a call to
339 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
340 records point to non-existent hosts", "retry timeout exceeded", and
341 "retry time not reached for any host after a long failure period".
343 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
344 experimental DomainKeys support:
346 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
347 (2) On an error such as an illegally used "control", the wrong name for
348 the control was given.
350 These problems did NOT occur unless DomainKeys support was compiled.
352 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
354 PH/24 Added ${if match_ip condition.
356 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
357 databases so that it will be absolutely obvious if a crash occurs in the
358 DB library. This is a regular occurrence (often caused by mis-matched
361 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
362 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
363 picked up on a user's system that detects such things. There doesn't seem
364 to be a gcc warning option for this - only an attribute that has to be
365 put on the function's prototype. It seems that in Fedora Core 4 they have
366 set this on a number of new functions. No doubt there will be more in due
369 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
370 specify a transport. However, if an address that was verified by such a
371 router was the subject of a callout, Exim crashed because it tried to
372 read the rcpt_include_affixes from the non-existent transport. Now it
373 just assumes that the setting of that option is false. This bug was
374 introduced by 4.51/PH/31.
376 PH/28 Changed -d+all to exclude +memory, because that information is very
377 rarely of interest, but it makes the output a lot bigger. People tend to
378 do -d+all out of habit.
380 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
381 code in os-type was giving problems when libc.so lives in lib64, like on
384 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
385 aren't the modern standard, and it seems that some systems' include files
386 don't always have them. Exim was already checking for some of the newer
387 ones like T_AAAA, and defining it itself. I've added checks for all the
388 record types that Exim uses.
390 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
391 not automatically generating a new one, as it is supposed to. This
392 prevented TLS from working. If the file did exist, but contained invalid
393 data, a new version was generated, as expected. It was only the case of a
394 non-existent file that was broken.
396 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
397 with a change in libdomainkeys > 0.64.
399 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
400 from DNS. If the selector record carries the flag, it now has
401 precedence over the domain-wide flag.
403 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
405 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
406 the use of a port name that isn't defined in /etc/services) occurred, the
407 message was deferred as in a normal delivery, and thus remained on the
408 spool, instead of being failed because of the mua_wrapper setting. This
409 is now fixed, and I tidied up some of the mua_wrapper messages at the
412 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
413 the messages in a hash of arrays rather than using individual hashes.
414 This is a bit cleaner and results in dramatic memory savings, albeit
415 at a slight CPU cost.
417 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
418 as requested by Marc Sherman.
420 SC/10 Eximstats: added histograms for user specified patterns as requested
423 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
425 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
426 fopen() in the content-scanning modules that did not already have it.
432 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
433 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
435 TK/02 Fix ACL "control" statement not being available in MIME ACL.
437 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
439 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
440 to test Sieve filters that use "vacation".
442 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
443 that changes the way the GnuTLS parameters are stored in the cache file.
444 The new format can be generated externally. For backward compatibility,
445 if the data in the cache doesn't make sense, Exim assumes it has read an
446 old-format file, and it generates new data and writes a new file. This
447 means that you can't go back to an older release without removing the
450 PH/03 A redirect router that has both "unseen" and "one_time" set does not
451 work if there are any delivery delays because "one_time" forces the
452 parent to be marked "delivered", so its unseen clone is never tried
453 again. For this reason, Exim now forbids the simultaneous setting of
456 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
457 redirected to themselves ("homonym" addresses). Read the long ChangeLog
458 entry if you want to know the details. The fix, however, neglected to
459 consider the case when local delivery batching is involved. The test for
460 "previously delivered" was not happening when checking to see if an
461 address could be batched with a previous (undelivered) one; under
462 certain circumstances this could lead to multiple deliveries to the same
465 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
466 in its include files, and this causes problems building Exim.
468 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
469 header_syntax) but Exim was just ignoring anything given after a slash.
470 In particular, this caused confusion with an attempt to use "verify =
471 reverse_host_lookup/defer_ok". An error is now given when options are
472 supplied for verify items that do not have them. (Maybe reverse_host_
473 lookup should have a defer_ok option, but that's a different point.)
475 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
476 defined by RFC 821) to 2048, because there were problems with some AUTH
477 commands, and RFC 1869 says the size should be increased for extended
478 SMTP commands that take arguments.
480 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
483 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
484 "unknown" error; now it says that the functionality isn't in the binary.
486 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
487 an address' error message when a string expansion fails (syntax or
488 whatever). Otherwise the password may appear in the log. Following change
489 PH/42 below, there is no longer a chance of it appearing in a bounce
492 PH/11 Installed exipick version 20050225.0 from John Jetmore.
494 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
495 of its addresses was ever tried. (Bugzilla bug #2.)
497 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
498 the result incorrectly in the debug output. (It correctly added a newline
499 to what was transported.)
501 TF/01 Added $received_time.
503 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
504 commented out examples of how to interface to a virus scanner and to
505 SpamAssassin. Also added commented examples of av_scanner and
506 spamd_address settings.
508 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
509 and controls are allowed in which ACLs. There were a couple of minor
510 errors. Some of the entries in the conditions table (which is a table of
511 where they are NOT allowed) were getting very unwieldy; rewrote them as a
512 negation of where the condition IS allowed.
514 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
516 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
517 header file does not have a version number, so I've had to invent a new
518 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
519 API. The code is untested by me (my Linux distribution still has 0.3.2 of
520 radiusclient), but it was contributed by a Radius user.
522 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
523 files or directories, for OpenSSL.
525 PH/19 When an Exim process that is running as root has to create an Exim log
526 file, it does so in a subprocess that runs as exim:exim so as to get the
527 ownership right at creation (otherwise, other Exim processes might see
528 the file with the wrong ownership). There was no test for failure of this
529 fork() call, which would lead to the process getting stuck as it waited
530 for a non-existent subprocess. Forks do occasionally fail when resources
531 run out. I reviewed all the other calls to fork(); they all seem to check
534 PH/20 When checking for unexpected SMTP input at connect time (before writing
535 the banner), Exim was not dealing correctly with a non-positive return
536 from the read() function. If the client had disconnected by this time,
537 the result was a log entry for a synchronization error with an empty
538 string after "input=" when read() returned zero. If read() returned -1
539 (an event I could not check), uninitialized data bytes were printed.
540 There were reports of junk text (parts of files, etc) appearing after
543 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
545 PH/22 Added support for macro redefinition, and (re)definition in between
546 driver and ACL definitions.
548 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
549 forgetting to use the resulting value; it was using the unexpanded value.
551 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
552 hadn't been configured. The fix is from Juergen Kreileder, who
553 understands it better than I do:
555 "Here's what I see happening with three configured cyrus_sasl
556 authenticators configured (plain, login, cram-md5):
558 On startup auth_cyrus_sasl_init() gets called for each of these.
559 This means three calls to sasl_listmech() without a specified mech_list.
560 => SASL tests which mechs of all available mechs actually work
561 => three warnings about OTP not working
562 => the returned list contains: plain, login, cram-md5, digest-md5, ...
564 With the patch, sasl_listmech() also gets called three times. But now
565 SASL's mech_list option is set to the server_mech specified in the the
566 authenticator. Or in other words, the answer from sasl_listmech()
567 gets limited to just the mech you're testing for (which is different
569 => the return list contains just 'plain' or 'login', 'cram-md5' or
570 nothing depending on the value of ob->server_mech.
572 I've just tested the patch: Authentication still works fine,
573 unavailable mechs specified in the exim configuration are still
574 caught, and the auth.log warnings about OTP are gone."
576 PH/25 When debugging is enabled, the contents of the command line are added
577 to the debugging output, even when log_selector=+arguments is not
580 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
581 answer is "GNU", and only if the return is "GNU/something" is the answer
584 PH/27 $acl_verify_message is now set immediately after the failure of a
585 verification in an ACL, and so is available in subsequent modifiers. In
586 particular, the message can be preserved by coding like this:
588 warn !verify = sender
589 set acl_m0 = $acl_verify_message
591 Previously, $acl_verify_message was set only while expanding "message"
592 and "log_message" when a very denied access.
594 PH/28 Modified OS/os.c-Linux with
596 -#ifndef OS_LOAD_AVERAGE
597 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
599 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
600 nomenclature these days.)
602 PH/29 Installed patch from the Sieve maintainer that adds the options
603 sieve_useraddress and sieve_subaddress to the redirect router.
605 PH/30 In these circumstances:
606 . Two addresses routed to the same list of hosts;
607 . First host does not offer TLS;
608 . First host accepts first address;
609 . First host gives temporary error to second address;
610 . Second host offers TLS and a TLS session is established;
611 . Second host accepts second address.
612 Exim incorrectly logged both deliveries with the TLS parameters (cipher
613 and peerdn, if requested) that were in fact used only for the second
616 PH/31 When doing a callout as part of verifying an address, Exim was not paying
617 attention to any local part prefix or suffix that was matched by the
618 router that accepted the address. It now behaves in the same way as it
619 does for delivery: the affixes are removed from the local part unless
620 rcpt_include_affixes is set on the transport.
622 PH/32 Add the sender address, as F=<...>, to the log line when logging a
623 timeout during the DATA phase of an incoming message.
625 PH/33 Sieve envelope tests were broken for match types other than :is. I have
626 applied a patch sanctioned by the Sieve maintainer.
628 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
629 the uid or gid is negative. A case of a negative gid caused this to be
630 noticed. The fix allows for either to be negative.
632 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
633 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
636 PH/36 The change PH/12 above was broken. Fixed it.
638 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
639 the grounds that routing the same address twice would always produce the
640 same answer. This might have been true once, but it is certainly no
641 longer true now. Routing a child address may depend on the previous
642 routing that produced that child. Some complicated redirection strategies
643 went wrong when messages had multiple recipients, and made Exim's
644 behaviour dependent on the order in which the addresses were given.
646 I have moved the duplicate checking until after the routing is complete.
647 Exim scans the addresses that are assigned to local and remote
648 transports, and removes any duplicates. This means that more work will be
649 done, as duplicates will always all be routed, but duplicates are
650 presumably rare, so I don't expect this is of any significance.
652 For deliveries to pipes, files, and autoreplies, the duplicate checking
653 still happens during the routing process, since they are not going to be
656 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
657 It corrects a timeout issue with spamd. This is Ian's comment: "The
658 background is that sometimes spamd either never reads data from a
659 connection it has accepted, or it never writes response data. The exiscan
660 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
661 blindly assumes that writes won't block so it may never time out."
663 PH/39 Allow G after quota size as well as K and M.
665 PH/40 The value set for $authenticated_id in an authenticator may not contain
666 binary zeroes or newlines because the value is written to log lines and
667 to spool files. There was no check on this. Now the value is run through
668 the string_printing() function so that such characters are converted to
669 printable escape sequences.
671 PH/41 $message_linecount is a new variable that contains the total number of
672 lines in the message. Compare $body_linecount, which is the count for the
675 PH/42 Exim no longer gives details of delivery errors for specific addresses in
676 bounce and delay warning messages, except in certain special cases, which
679 (a) An SMTP error message from a remote host;
680 (b) A message specified in a :fail: redirection;
681 (c) A message specified in a "fail" command in a system filter;
682 (d) A message specified in a FAIL return from the queryprogram router;
683 (e) A message specified by the cannot_route_message router option.
685 In these cases only, Exim does include the error details in bounce and
686 warning messages. There are also a few cases where bland messages such
687 as "unrouteable address" or "local delivery error" are given.
689 PH/43 $value is now also set for the "else" part of a ${run expansion.
691 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
692 being worked on, but at least Exim now implements the latest version to
695 PH/45 In a pipe transport, although a timeout while waiting for the pipe
696 process to complete was treated as a delivery failure, a timeout while
697 writing the message to the pipe was logged, but erroneously treated as a
698 successful delivery. Such timeouts include transport filter timeouts. For
699 consistency with the overall process timeout, these timeouts are now
700 treated as errors, giving rise to delivery failures by default. However,
701 there is now a new Boolean option for the pipe transport called
702 timeout_defer, which, if set TRUE, converts the failures into defers for
703 both kinds of timeout. A transport filter timeout is now identified in
706 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
707 systems where "make" and "gmake" are different, calling "gmake" at top
708 level broke things. I've arranged for the value of $(MAKE) to be passed
709 from the Makefile to this script so that it can call the same version of
713 A note about Exim versions 4.44 and 4.50
714 ----------------------------------------
716 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
717 changes of various kinds. As a consequence, a big documentation update was
718 needed. This delayed the release for rather longer than seemed good, especially
719 in the light of a couple of (minor) security issues. Therefore, the changes
720 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
721 release. So 4.44 and 4.50 are in effect two different branches that both start
724 I have left the 4.50 change log unchanged; it contains all the changes since
725 4.43. The change log for 4.44 is below; many of its items are identical to
726 those for 4.50. This seems to be the most sensible way to preserve the
727 historical information.
733 1. Minor wording change to the doc/README.SIEVE file.
735 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
736 computation of the current number of files was incorrect.
738 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
739 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
742 4. Give more explanation in the error message when the command for a transport
743 filter fails to execute.
745 5. There are several places where Exim runs a non-Exim command in a
746 subprocess. The SIGUSR1 signal should be disabled for these processes. This
747 was being done only for the command run by the queryprogram router. It is
748 now done for all such subprocesses. The other cases are: ${run, transport
749 filters, and the commands run by the lmtp and pipe transports.
751 6. Added CONFIGURE_GROUP build-time option.
753 7. Some older OS have a limit of 256 on the maximum number of file
754 descriptors. Exim was using setrlimit() to set 1000 as a large value
755 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
756 systems. I've change it so that if it can't get 1000, it tries for 256.
758 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
759 was an oversight, and furthermore, ever since the addition of extra
760 controls (e.g. 4.43/32), the checks on when to allow different forms of
761 "control" were broken. There should now be diagnostics for all cases when a
762 control that does not make sense is encountered.
764 9. Added the /retain_sender option to "control=submission".
766 10. $recipients is now available in the predata ACL (oversight).
768 11. Tidy the search cache before the fork to do a delivery from a message
769 received from the command line. Otherwise the child will trigger a lookup
770 failure and thereby defer the delivery if it tries to use (for example) a
771 cached ldap connection that the parent has called unbind on.
773 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
774 of $address_data from the recipient verification was clobbered by the
777 13. The value of address_data from a sender verification is now available in
778 $sender_address_data in subsequent conditions in the ACL statement.
780 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
782 15. Added a new option "connect=<time>" to callout options, to set a different
785 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
786 was its contents. (It was OK if the option was not defined at all.)
788 17. A "Completed" log line is now written for messages that are removed from
789 the spool by the -Mrm option.
791 18. New variables $sender_verify_failure and $recipient_verify_failure contain
792 information about exactly what failed.
794 19. Added -dd to debug only the daemon process.
796 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
797 handles timeouts, both on the server side and network timeouts. Renamed the
798 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
800 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
801 It is now set to "smtps".
803 22. $host_address is now set to the target address during the checking of
806 23. When checking ignore_target_hosts for an ipliteral router, no host name was
807 being passed; this would have caused $sender_host_name to have been used if
808 matching the list had actually called for a host name (not very likely,
809 since this list is usually IP addresses). A host name is now passed as
812 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
813 code that specifies a non-restarting handler (typically sigaction() in
814 modern systems) in an attempt to fix a rare and obscure crash bug.
816 25. Narrowed the window for a race in the daemon that could cause it to ignore
817 SIGCHLD signals. This is not a major problem, because they are used only to
818 wake it up if nothing else does.
820 26. A malformed maildirsize file could cause Exim to calculate negative values
821 for the mailbox size or file count. Odd effects could occur as a result.
822 The maildirsize information is now recalculated if the size or filecount
825 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
826 support for a long time. Removed HAVE_SYS_VFS_H.
828 28. Installed the latest version of exipick from John Jetmore.
830 29. In an address list, if the pattern was not a regular expression, an empty
831 subject address (from a bounce message) matched only if the pattern was an
832 empty string. Non-empty patterns were not even tested. This was the wrong
833 because it is perfectly reasonable to use an empty address as part of a
834 database query. An empty address is now tested by patterns that are
835 lookups. However, all the other forms of pattern expect the subject to
836 contain a local part and a domain, and therefore, for them, an empty
837 address still always fails if the pattern is not itself empty.
839 30. Exim went into a mad DNS loop when attempting to do a callout where the
840 host was specified on an smtp transport, and looking it up yielded more
843 31. Re-factored the code for checking spool and log partition space into a
844 function that finds that data and another that does the check. The former
845 is then used to implement four new variables: $spool_space, $log_space,
846 $spool_inodes, and $log_inodes.
848 32. The RFC2047 encoding function was originally intended for short strings
849 such as real names; it was not keeping to the 75-character limit for
850 encoded words that the RFC imposes. It now respects the limit, and
851 generates multiple encoded words if necessary. To be on the safe side, I
852 have increased the buffer size for the ${rfc2047: expansion operator from
855 33. It is now permitted to omit both strings after an "if" condition; if the
856 condition is true, the result is "true". As before, when the second string
857 is omitted, a false condition yields an empty string. This makes it less
858 cumbersome to write custom ACL and router conditions.
860 34. Failure to deliver a bounce message always caused it to be frozen, even if
861 there was an errors_to setting on the router. The errors_to setting is now
864 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
865 canonical form (fully expanded) before being placed in
866 $sender_host_address.
868 36. The table in the code that translates DNS record types into text (T_A to
869 "A" for instance) was missing entries for NS and CNAME. It is just possible
870 that this could have caused confusion if both these types were looked up
871 for the same domain, because the text type is used as part of Exim's
872 per-process caching. But the chance of anyone hitting this buglet seems
875 37. The dnsdb lookup has been extended in a number of ways.
877 (1) There is a new type, "zns", which walks up the domain tree until it
878 finds some nameserver records. It should be used with care.
880 (2) There is a new type, "mxh", which is like "mx" except that it returns
881 just the host names, not the priorities.
883 (3) It is now possible to give a list of domains (or IP addresses) to be
884 looked up. The behaviour when one of the lookups defers can be
885 controlled by a keyword.
887 (4) It is now possible to specify the separator character for use when
888 multiple records are returned.
890 38. The dnslists ACL condition has been extended: it is now possible to supply
891 a list of IP addresses and/or domains to be looked up in a particular DNS
894 39. Added log_selector=+queue_time_overall.
896 40. When running the queue in the test harness, wait just a tad after forking a
897 delivery process, to get repeatability of debugging output.
899 41. Include certificate and key file names in error message when GnuTLS fails
900 to set them up, because the GnuTLS error message doesn't include the name
901 of the failing file when there is a problem reading it.
903 42. Allow both -bf and -bF in the same test run.
905 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
907 44. The "Exiscan patch" is now merged into the mainline Exim source.
909 45. Sometimes the final signoff response after QUIT could fail to get
910 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
911 < 0 before doing a fflush(). This bug looks as though it goes back to the
912 introduction of TLS in release 3.20, but "sometimes" must have been rare
913 because the tests only now provoked it.
915 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
916 (this can affect the format of dates).
918 47. exim_tidydb, when checking for the continued existence of a message for
919 which it has found a message-specific retry record, was not finding
920 messages that were in split spool directories. Consequently, it was
921 deleting retry records that should have stayed in existence.
923 48. Steve fixed some bugs in eximstats.
925 49. The SPA authentication driver was not abandoning authentication and moving
926 on to the next authenticator when an expansion was forced to fail,
927 contradicting the general specification for all authenticators. Instead it
928 was generating a temporary error. It now behaves as specified.
930 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
931 (the order specifies the preference for clients). The order is now AES256,
932 AES128, 3DES, ARCFOUR128.
934 51. Small patch to Sieve code - explicitly set From: when generating an
937 52. Exim crashed if a remote delivery caused a very long error message to be
938 recorded - for instance if somebody sent an entire SpamAssassin report back
939 as a large number of 550 error lines. This bug was coincidentally fixed by
940 increasing the size of one of Exim's internal buffers (big_buffer) that
941 happened as part of the Exiscan merge. However, to be on the safe side, I
942 have made the code more robust (and fixed the comments that describe what
945 53. Now that there can be additional text after "Completed" in log lines (if
946 the queue_time_overall log selector is set), a one-byte patch to exigrep
947 was needed to allow it to recognize "Completed" as not the last thing in
950 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
951 patch that reportedly fixes this has been added. I am not expert enough to
952 create a test for it. This is what the patch creator wrote:
954 "I found a little strange behaviour of ldap code when working with
955 Windows 2003 AD Domain, where users was placed in more than one
956 Organization Units. When I tried to give exim partial DN, the exit code
957 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
958 But simultaneously result of request was absolutely normal ldap result,
959 so I produce this patch..."
961 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
962 REFERENCE, so I have modified the code to exclude the patch when that macro
965 55. Some experimental protocols are using DNS PTR records for new purposes. The
966 keys for these records are domain names, not reversed IP addresses. The
967 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
968 leaves it alone. Component reversal etc. now happens only for IP addresses.
971 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
973 57. Double the size of the debug message buffer (to 2048) so that more of very
974 long debug lines gets shown.
976 58. The exicyclog utility now does better if the number of log files to keep
977 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
979 59. Two changes related to the smtp_active_hostname option:
981 (1) $smtp_active_hostname is now available as a variable.
982 (2) The default for smtp_banner uses $smtp_active_hostname instead
983 of $primary_hostname.
985 60. The host_aton() function is supposed to be passed a string that is known
986 to be a valid IP address. However, in the case of IPv6 addresses, it was
987 not checking this. This is a hostage to fortune. Exim now panics and dies
988 if the condition is not met. A case was found where this could be provoked
989 from a dnsdb PTR lookup with an IPv6 address that had more than 8
990 components; fortuitously, this particular loophole had already been fixed
991 by change 4.50/55 above.
993 If there are any other similar loopholes, the new check in host_aton()
994 itself should stop them being exploited. The report I received stated that
995 data on the command line could provoke the exploit when Exim was running as
996 exim, but did not say which command line option was involved. All I could
997 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1001 61. There was a buffer overflow vulnerability in the SPA authentication code
1002 (which came originally from the Samba project). I have added a test to the
1003 spa_base64_to_bits() function which I hope fixes it.
1006 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1007 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1009 63. The daemon start-up calls getloadavg() while still root for those OS that
1010 need the first call to be done as root, but it missed one case: when
1011 deliver_queue_load_max is set with deliver_drop_privilege. This is
1012 necessary for the benefit of the queue runner, because there is no re-exec
1013 when deliver_drop_privilege is set.
1015 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1016 This has been fixed.
1018 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1019 was in use, was not putting the data itself into the right store pool;
1020 consequently, it could be overwritten for a subsequent message in the same
1021 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1024 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1026 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1027 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1028 the function were treating the return as a boolean value, which happened to
1029 work because 0=false and not-0=true, but is not correct code.
1031 68. The host_aton() function was not handling scoped IPv6 addresses (those
1032 with, for example, "%eth0" on the end) correctly.
1034 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1035 negated items (that is, ~something) in unsigned ints. Some compilers
1036 apparently mutter when there is no cast.
1038 70. If an address verification called from an ACL failed, and did not produce a
1039 user-specific message (i.e. there was only a "system" message), nothing was
1040 put in $acl_verify_message. In this situation, it now puts the system
1043 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1044 session; change 4.31/43 added the unwanted input to the log line - except
1045 that it did not do this in the start of session case. It now does.
1047 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1048 This is wrong and can cause the other end to generate a synchronization
1049 error if it is another Exim or anything else that does the synchronization
1050 check. A QUIT command is no longer sent after a timeout.
1052 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1053 during host lookups.
1055 74. The defer_ok option of callout verification was not working if it was used
1056 when verifying addresses in header lines, that is, for this case:
1058 verify = header_sender/callout=defer_ok
1060 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1061 those file descriptors could be used for SMTP connections. If anything
1062 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1063 could be sent to the SMTP client, causing chaos. The daemon now opens
1064 stdin, stdout, and stderr to /dev/null when it puts itself into the
1067 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1068 log by default. The user can override this with suitable Perl magic.
1070 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1071 the log message when discard triggers, was not working for the DATA ACL or
1072 for the non-SMTP ACL.
1074 78. Error message wording change in sieve.c.
1076 79. If smtp_accept_max_per_host was set, the number of connections could be
1077 restricted to fewer than expected, because the daemon was trying to set up
1078 a new connection before checking whether the processes handling previous
1079 connections had finished. The check for completed processes is now done
1080 earlier. On busy systems, this bug wouldn't be noticed because something
1081 else would have woken the daemon, and it would have reaped the completed
1084 80. If a message was submitted locally by a user whose login name contained one
1085 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1086 It caused a spool format error. I have fixed the spool reading code. A
1087 related problem was that the "from" clause in the Received: line became
1088 illegal because of the space(s). It is now covered by ${quote_local_part.
1090 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1093 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1095 83. Patch from Sieve maintainer to fix unterminated string problem in
1096 "vacation" handling.
1098 84. Some minor changes to the Linux configuration files to help with other
1099 OS variants using glibc.
1101 85. One more patch for Sieve to update vacation handling to latest spec.
1104 ----------------------------------------------------
1105 See the note above about the 4.44 and 4.50 releases.
1106 ----------------------------------------------------
1112 1. Change 4.43/35 introduced a bug that caused file counts to be
1113 incorrectly computed when quota_filecount was set in an appendfile
1116 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1117 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1120 3. Give more explanation in the error message when the command for a transport
1121 filter fails to execute.
1123 4. There are several places where Exim runs a non-Exim command in a
1124 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1125 was being done only for the command run by the queryprogram router. It is
1126 now done for all such subprocesses. The other cases are: ${run, transport
1127 filters, and the commands run by the lmtp and pipe transports.
1129 5. Some older OS have a limit of 256 on the maximum number of file
1130 descriptors. Exim was using setrlimit() to set 1000 as a large value
1131 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1132 systems. I've change it so that if it can't get 1000, it tries for 256.
1134 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1135 was an oversight, and furthermore, ever since the addition of extra
1136 controls (e.g. 4.43/32), the checks on when to allow different forms of
1137 "control" were broken. There should now be diagnostics for all cases when a
1138 control that does not make sense is encountered.
1140 7. $recipients is now available in the predata ACL (oversight).
1142 8. Tidy the search cache before the fork to do a delivery from a message
1143 received from the command line. Otherwise the child will trigger a lookup
1144 failure and thereby defer the delivery if it tries to use (for example) a
1145 cached ldap connection that the parent has called unbind on.
1147 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1148 of $address_data from the recipient verification was clobbered by the
1149 sender verification.
1151 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1152 was its contents. (It was OK if the option was not defined at all.)
1154 11. A "Completed" log line is now written for messages that are removed from
1155 the spool by the -Mrm option.
1157 12. $host_address is now set to the target address during the checking of
1158 ignore_target_hosts.
1160 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1161 being passed; this would have caused $sender_host_name to have been used if
1162 matching the list had actually called for a host name (not very likely,
1163 since this list is usually IP addresses). A host name is now passed as
1166 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1167 code that specifies a non-restarting handler (typically sigaction() in
1168 modern systems) in an attempt to fix a rare and obscure crash bug.
1170 15. Narrowed the window for a race in the daemon that could cause it to ignore
1171 SIGCHLD signals. This is not a major problem, because they are used only to
1172 wake it up if nothing else does.
1174 16. A malformed maildirsize file could cause Exim to calculate negative values
1175 for the mailbox size or file count. Odd effects could occur as a result.
1176 The maildirsize information is now recalculated if the size or filecount
1179 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1180 support for a long time. Removed HAVE_SYS_VFS_H.
1182 18. Updated exipick to current release from John Jetmore.
1184 19. Allow an empty sender to be matched against a lookup in an address list.
1185 Previously the only cases considered were a regular expression, or an
1188 20. Exim went into a mad DNS lookup loop when doing a callout where the
1189 host was specified on the transport, if the DNS lookup yielded more than
1192 21. The RFC2047 encoding function was originally intended for short strings
1193 such as real names; it was not keeping to the 75-character limit for
1194 encoded words that the RFC imposes. It now respects the limit, and
1195 generates multiple encoded words if necessary. To be on the safe side, I
1196 have increased the buffer size for the ${rfc2047: expansion operator from
1199 22. Failure to deliver a bounce message always caused it to be frozen, even if
1200 there was an errors_to setting on the router. The errors_to setting is now
1203 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1204 canonical form (fully expanded) before being placed in
1205 $sender_host_address.
1207 24. Updated eximstats to version 1.33
1209 25. Include certificate and key file names in error message when GnuTLS fails
1210 to set them up, because the GnuTLS error message doesn't include the name
1211 of the failing file when there is a problem reading it.
1213 26. Expand error message when OpenSSL has problems setting up cert/key files.
1216 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1217 (this can affect the format of dates).
1219 28. exim_tidydb, when checking for the continued existence of a message for
1220 which it has found a message-specific retry record, was not finding
1221 messages that were in split spool directories. Consequently, it was
1222 deleting retry records that should have stayed in existence.
1224 29. eximstats updated to version 1.35
1225 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1226 1.35 - bugfix such that pie charts by volume are generated correctly
1228 30. The SPA authentication driver was not abandoning authentication and moving
1229 on to the next authenticator when an expansion was forced to fail,
1230 contradicting the general specification for all authenticators. Instead it
1231 was generating a temporary error. It now behaves as specified.
1233 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1234 (the order specifies the preference for clients). The order is now AES256,
1235 AES128, 3DES, ARCFOUR128.
1237 31. Small patch to Sieve code - explicitly set From: when generating an
1240 32. Exim crashed if a remote delivery caused a very long error message to be
1241 recorded - for instance if somebody sent an entire SpamAssassin report back
1242 as a large number of 550 error lines. This bug was coincidentally fixed by
1243 increasing the size of one of Exim's internal buffers (big_buffer) that
1244 happened as part of the Exiscan merge. However, to be on the safe side, I
1245 have made the code more robust (and fixed the comments that describe what
1248 33. Some experimental protocols are using DNS PTR records for new purposes. The
1249 keys for these records are domain names, not reversed IP addresses. The
1250 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1251 leaves it alone. Component reversal etc. now happens only for IP addresses.
1254 34. The host_aton() function is supposed to be passed a string that is known
1255 to be a valid IP address. However, in the case of IPv6 addresses, it was
1256 not checking this. This is a hostage to fortune. Exim now panics and dies
1257 if the condition is not met. A case was found where this could be provoked
1258 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1259 components; fortuitously, this particular loophole had already been fixed
1260 by change 4.50/55 or 4.44/33 above.
1262 If there are any other similar loopholes, the new check in host_aton()
1263 itself should stop them being exploited. The report I received stated that
1264 data on the command line could provoke the exploit when Exim was running as
1265 exim, but did not say which command line option was involved. All I could
1266 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1267 running as the user.
1270 35. There was a buffer overflow vulnerability in the SPA authentication code
1271 (which came originally from the Samba project). I have added a test to the
1272 spa_base64_to_bits() function which I hope fixes it.
1275 36. The daemon start-up calls getloadavg() while still root for those OS that
1276 need the first call to be done as root, but it missed one case: when
1277 deliver_queue_load_max is set with deliver_drop_privilege. This is
1278 necessary for the benefit of the queue runner, because there is no re-exec
1279 when deliver_drop_privilege is set.
1281 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1282 was in use, was not putting the data itself into the right store pool;
1283 consequently, it could be overwritten for a subsequent message in the same
1284 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1287 38. Sometimes the final signoff response after QUIT could fail to get
1288 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1289 < 0 before doing a fflush(). This bug looks as though it goes back to the
1290 introduction of TLS in release 3.20, but "sometimes" must have been rare
1291 because the tests only now provoked it.
1297 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1298 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1299 Now it returns an integer. A number of calls were still expecting a T/F
1300 return. Fortuitously, in all cases, the tests worked in OK situations,
1301 which is the norm. However, things would have gone wrong on any write
1302 failures on the smtp file descriptor. This function is used when sending
1303 messages over SMTP and also when doing verify callouts.
1305 2. When Exim is called to do synchronous delivery of a locally submitted
1306 message (the -odf or -odi options), it no longer closes stderr before doing
1309 3. Implemented the mua_wrapper option.
1311 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1313 5. Implemented the functions header_remove(), header_testname(),
1314 header_add_at_position(), and receive_remove_recipient(), and exported them
1317 6. If an ACL "warn" statement specified the addition of headers, Exim already
1318 inserted X-ACL-Warn: at the start if there was no header name. However, it
1319 was not making this test for the second and subsequent header lines if
1320 there were newlines in the string. This meant that an invalid header could
1321 be inserted if Exim was badly configured.
1323 7. Allow an ACL "warn" statement to add header lines at the start or after all
1324 the Received: headers, as well as at the end.
1326 8. Added the rcpt_4xx retry error code.
1328 9. Added postmaster_mailfrom=xxx to callout verification option.
1330 10. Added mailfrom=xxxx to the callout verification option, for verify=
1333 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1334 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1336 12. Inserted some casts to stop certain compilers warning when using pointer
1337 differences as field lengths or precisions in printf-type calls (mostly
1338 affecting debugging statements).
1340 13. Added optional readline() support for -be (dynamically loaded).
1342 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1343 same clock tick as a message's arrival, so that its received time was the
1344 same as the "first fail" time on the retry record, and that message
1345 remained on the queue past the ultimate address timeout, every queue runner
1346 would try a delivery (because it was past the ultimate address timeout) but
1347 after another failure, the ultimate address timeout, which should have then
1348 bounced the address, did not kick in. This was a "< instead of <=" error;
1349 in most cases the first failure would have been in the next clock tick
1350 after the received time, and all would be well.
1352 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1353 being recognized when the domain list was tested by the match_domain
1354 condition in an expansion string.
1356 16. Added the ${str2b64: operator.
1358 17. Exim was always calling setrlimit() to set a large limit for the number of
1359 processes, without checking whether the existing limit was already
1360 adequate. (It did check for the limit on file descriptors.) Furthermore,
1361 errors from getrlimit() and setrlimit() were being ignored. Now they are
1362 logged to the main and panic logs, but Exim does carry on, to try to do its
1363 job under whatever limits there are.
1365 18. Imported PCRE 5.0.
1367 19. Trivial typo in log message " temporarily refused connection" (the leading
1370 20. If the log selector return_path_on_delivery was set and an address was
1371 redirected to /dev/null, the delivery process crashed because it assumed
1372 that a return path would always be set for a "successful" delivery. In this
1373 case, the whole delivery is bypassed as an optimization, and therefore no
1376 21. Internal re-arrangement: the function for sending a challenge and reading
1377 a response while authentication was assuming a zero-terminated challenge
1378 string. It's now changed to take a pointer and a length, to allow for
1379 binary data in such strings.
1381 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1383 23. Exim was not respecting finduser_retries when seeking the login of the
1384 uid under which it was called; it was always trying 10 times. (The default
1385 setting of finduser_retries is zero.) Also, it was sleeping after the final
1386 failure, which is pointless.
1388 24. Implemented tls_on_connect_ports.
1390 25. Implemented acl_smtp_predata.
1392 26. If the domain in control=submission is set empty, Exim assumes that the
1393 authenticated id is a complete email address when it generates From: or
1394 Sender: header lines.
1396 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1397 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1398 chown and chgrp in /bin and hostname in /usr/bin.
1400 28. Exim was keeping the "process log" file open after each use, just as it
1401 does for the main log. This opens the possibility of it remaining open for
1402 long periods when the USR1 signal hits a daemon. Occasional processlog
1403 errors were reported, that could have been caused by this. Anyway, it seems
1404 much more sensible not to leave this file open at all, so that is what now
1407 29. The long-running daemon process does not normally write to the log once it
1408 has entered its main loop, and it closes the log before doing so. This is
1409 so that log files can straightforwardly be renamed and moved. However,
1410 there are a couple of unusual error situations where the daemon does write
1411 log entries, and I had neglected to close the log afterwards.
1413 30. The text of an SMTP error response that was received during a remote
1414 delivery was being truncated at 512 bytes. This is too short for some of
1415 the long messages that one sometimes sees. I've increased the limit to
1418 31. It is now possible to make retry rules that apply only when a message has a
1419 specific sender, in particular, an empty sender.
1421 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1422 it possible to be selective about when SMTP synchronization is enforced.
1424 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1426 32. Implemented hosts_connection_nolog.
1428 33. Added an ACL for QUIT.
1430 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1433 35. Added mailbox_size and mailbox_filecount to appendfile.
1435 36. Added control = no_multiline_responses to ACLs.
1437 37. There was a bug in the logic of the code that waits for the clock to tick
1438 in the case where the clock went backwards by a substantial amount such
1439 that the microsecond fraction of "now" was more than the microsecond
1440 fraction of "then" (but the whole seconds number was less).
1442 38. Added support for the libradius Radius client library this is found on
1443 FreeBSD (previously only the radiusclient library was supported).
1449 1. When certain lookups returned multiple values in the form name=value, the
1450 quoting of the values was not always being done properly. Specifically:
1451 (a) If the value started with a double quote, but contained no whitespace,
1453 (b) If the value contained whitespace other than a space character (i.e.
1454 tabs or newlines or carriage returns) it was not quoted.
1455 This fix has been applied to the mysql and pgsql lookups by writing a
1456 separate quoting function and calling it from the lookup code. The fix
1457 should probably also be applied to nisplus, ibase and oracle lookups, but
1458 since I cannot test any of those, I have not disturbed their existing code.
1460 2. A hit in the callout cache for a specific address caused a log line with no
1461 reason for rejecting RCPT. Now it says "Previous (cached) callout
1462 verification failure".
1464 3. There was an off-by-one bug in the queryprogram router. An over-long
1465 return line was truncated at 256 instead of 255 characters, thereby
1466 overflowing its buffer with the terminating zero. As well as fixing this, I
1467 have increased the buffer size to 1024 (and made a note to document this).
1469 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1470 when Exim is waiting for an SMTP response from a remote server, Exim
1471 restarts its select() call on the socket, thereby resetting its timeout.
1472 This is not a problem when such interrupts are rare. Somebody set up a cron
1473 job to run exiwhat every 2 minutes, which is less than the normal select()
1474 timeout (5 or 10 minutes). This meant that the select() timeout never
1475 kicked in because it was always reset. I have fixed this by comparing the
1476 time when an interrupt arrives with the time at the start of the first call
1477 to select(). If more time than the timeout has elapsed, the interrupt is
1478 treated as a timeout.
1480 5. Some internal re-factoring in preparation for the addition of Sieve
1481 extensions (by MH). In particular, the "personal" test is moved to a
1482 separate function, and given an option for scanning Cc: and Bcc: (which is
1483 not set for Exim filters).
1485 6. When Exim created an email address using the login of the caller as the
1486 local part (e.g. when creating a From: or Sender: header line), it was not
1487 quoting the local part when it contained special characters such as @.
1489 7. Installed new OpenBSD configuration files.
1491 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1492 try to make them clearer.
1494 9. Callout options, other than the timeout value, were being ignored when
1495 verifying sender addresses in header lines. For example, when using
1497 verify = header_sender/callout=no_cache
1499 the cache was (incorrectly) being used.
1501 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1502 only the Cygwin environment.
1504 11. When return_path_on_delivery was set as a log selector, if different remote
1505 addresses in the same message used different return paths and parallel
1506 remote delivery occurred, the wrong values would sometimes be logged.
1507 (Whenever a remote delivery process finished, the return path value from
1508 the most recently started remote delivery process was logged.)
1510 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1511 lines when AUTH and/or TLS are in use. This is the "received protocol"
1512 field. Exim used to use "asmtp" for authenticated SMTP, without any
1513 indication (in the protocol name) for TLS use. Now it follows the RFC and
1514 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1515 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1516 names appear in log lines as well as in Received: header lines.
1518 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1519 extensions, and comparison tests, and to fix some bugs.
1521 14. Changes to the "personal" filter test:
1523 (1) The test was buggy in that it was just doing the equivalent of
1524 "contains" tests on header lines. For example, if a user's address was
1525 anne@some.where, the "personal" test would incorrectly be true for
1527 To: susanne@some.where
1529 This test is now done by extracting each address from the header in turn,
1530 and checking the entire address. Other tests that are part of "personal"
1531 are now done using regular expressions (for example, to check local parts
1532 of addresses in From: header lines).
1534 (2) The list of non-personal local parts in From: addresses has been
1535 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1536 taken from the Sieve specification recommendations.
1538 (3) If the message contains any header line starting with "List-" it is
1539 treated as non-personal.
1541 (4) The test for "circular" in the Subject: header line has been removed
1542 because it now seems ill-conceived.
1544 15. Minor typos in src/EDITME comments corrected.
1546 16. Installed latest exipick from John Jetmore.
1548 17. If headers_add on a router specified a text string that was too long for
1549 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1550 of string_sprintf() is now avoided.
1552 18. $message_body_size was not set (it was always zero) when running the DATA
1553 ACL and the local_scan() function.
1555 19. For the "mail" command in an Exim filter, no default was being set for
1556 the once_repeat time, causing a random time value to be used if "once" was
1557 specified. (If the value happened to be <= 0, no repeat happened.) The
1558 default is now 0s, meaning "never repeat". The "vacation" command was OK
1559 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1560 (I found it when inspecting the code).
1562 20. There is now an overall timeout for performing a callout verification. It
1563 defaults to 4 times the callout timeout, which applies to individual SMTP
1564 commands during the callout. The overall timeout applies when there is more
1565 than one host that can be tried. The timeout is checked before trying the
1566 next host. This prevents very long delays if there are a large number of
1567 hosts and all are timing out (e.g. when the network connections are timing
1568 out). The value of the overall timeout can be changed by specifying an
1569 additional sub-option for "callout", called "maxwait". For example:
1571 verify = sender/callout=5s,maxwait=20s
1573 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1574 to the end before writing, but this should make it even safer).
1576 22. Exim was forgetting that it had advertised PIPELINING for the second and
1577 subsequent messages on an SMTP connection. It was also not resetting its
1578 memory on STARTTLS and an internal HELO.
1580 23. When Exim logs an SMTP synchronization error within a session, it now
1581 records whether PIPELINING has been advertised or not.
1583 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1584 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1585 rather than long int.
1587 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1589 26. Added the never_mail option to autoreply.
1595 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1596 crash if the getsockname() call failed; this can happen if a connection is
1597 closed very soon after it is established. The problem was simply in the
1598 order in which certain operations were done, causing Exim to try to write
1599 to the SMTP stream before it had set up the file descriptor. The bug has
1600 been fixed by making things happen in the correct order.
1606 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1607 before the connection was closed, thus losing the rejection response.
1609 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1610 some early Solaris releases, but causes trouble in current releases where
1611 socklen_t is defined.
1613 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1616 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1619 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1620 that releasing the top of it at the end releases what was used for sub-
1621 expansions (unless the block got too big). However, discard this block if
1622 the first thing is a variable or header, so that we can use its block when
1623 it is dynamic (useful for very large $message_headers, for example).
1625 6. Lookups now cache *every* query, not just the most recent. A new, separate
1626 store pool is used for this. It can be recovered when all lookup caches are
1627 flushed. Lookups now release memory at the end of their result strings.
1628 This has involved some general refactoring of the lookup sources.
1630 7. Some code has been added to the store_xxx() functions to reduce the amount
1631 of flapping under certain conditions.
1633 8. log_incoming_interface used to affect only the <= reception log lines. Now
1634 it causes the local interface and port to be added to several more SMTP log
1635 lines, for example "SMTP connection from", and rejection lines.
1637 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1639 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1641 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1642 could be overwritten at the end of the current message (or the start of a
1643 new message if it was set in a HELO ACL). The value is now preserved for
1644 the duration of the SMTP connection.
1646 12. If a transport had a headers_rewrite setting, and a matching header line
1647 contained an unqualified address, that address was qualified, even if it
1648 did not match any rewriting rules. The underlying bug was that the values
1649 of the flags that permit the existence of unqualified sender and recipient
1650 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1651 non-local messages, and by -bnq for local messages) were not being
1652 preserved with the message after it was received.
1654 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1655 "next input=" as part of the text comprising the host identity instead of
1656 the correct text. The code was using the same buffer for two different
1657 strings. However, depending on which order the printing function evaluated
1658 its arguments, the bug did not always show up. Under Linux, for example, my
1659 test suite worked just fine.
1661 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1662 doesn't work with some older versions of Perl. It has been changed to "my",
1663 which in any case is probably the better facility to use.
1665 15. A really picky compiler found some instances of statements for creating
1666 error messages that either had too many or two few arguments for the format
1669 16. The size of the buffer for calls to the DNS resolver has been increased
1670 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1671 for addresses that have a lot of PTR records. This alleviates a problem; it
1672 does not fully solve it.
1674 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1675 buffer now truncates the list and logs the incident, which is the same
1676 action as happens when Exim is looking up a host name and its aliases.
1677 Previously in this situation something unpredictable would happen;
1678 sometimes it was "internal error: store_reset failed".
1680 18. If a server dropped the connection unexpectedly when an Exim client was
1681 using GnuTLS and trying to read a response, the client delivery process
1682 crashed while trying to generate an error log message.
1684 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1685 string, for example:
1687 warn message = H1: something\nH2: something
1689 the text was added as a single header line from Exim's point of view
1690 though it ended up OK in the delivered message. However, searching for the
1691 second and subsequent header lines using $h_h2: did not work. This has been
1692 fixed. Similarly, if a system filter added multiple headers in this way,
1693 the routers could not see them.
1695 20. Expanded the error message when iplsearch is called with an invalid key to
1696 suggest using net-iplsearch in a host list.
1698 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1699 ACLs are no longer actually imposed (and a message to that effect is
1702 22. If a "gecos" field in a passwd entry contained escaped characters, in
1703 particular, if it contained a \" sequence, Exim got it wrong when building
1704 a From: or a Sender: header from that name. A second bug also caused
1705 incorrect handling when an unquoted " was present following a character
1706 that needed quoting.
1708 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1709 was not being matched caselessly.
1711 24. Arranged for all hyphens in the exim.8 source to be escaped with
1714 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1715 properly. Recipient callout cache records were still being keyed to include
1716 the sender, even when use_sender was set false. This led to far more
1717 callouts that were necessary. The sender is no longer included in the key
1718 when use_sender is false.
1720 26. Added "control = submission" modifier to ACLs.
1722 27. Added the ${base62d: operator to decode base 62 numbers.
1724 28. dnsdb lookups can now access SRV records.
1726 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1727 the configuration file.
1729 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1730 (-v) mode. This makes the output for a verbose queue run more intelligible.
1732 31. Added a use_postmaster feature to recipient callouts.
1734 32. Added the $body_zerocount variable, containing the number of binary zero
1735 bytes in the message body.
1737 33. The time of last modification of the "new" subdirectory is now used as the
1738 "mailbox time last read" when there is a quota error for a maildir
1741 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1743 35. Added +ignore_unknown as a special item in host lists.
1745 36. Code for decoding IPv6 addresses in host lists is now included, even if
1746 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1747 address was recognized as an IP address, but was then not correctly decoded
1748 into binary, causing unexpected and incorrect effects when compared with
1755 1. Very minor rewording of debugging text in manualroute to say "list of
1756 hosts" instead of "hostlist".
1758 2. If verify=header_syntax was set, and a header line with an unqualified
1759 address (no domain) and a large number of spaces between the end of the
1760 name and the colon was received, the reception process suffered a buffer
1761 overflow, and (when I tested it) crashed. This was caused by some obsolete
1762 code that should have been removed. The fix is to remove it!
1764 3. When running in the test harness, delay a bit after writing a bounce
1765 message to get a bit more predictability in the log output.
1767 4. Added a call to search_tidyup() just before forking a reception process. In
1768 theory, someone could use a lookup in the expansion of smtp_accept_max_
1769 per_host which, without the tidyup, could leave open a database connection.
1771 5. Added the variables $recipient_data and $sender_data which get set from a
1772 lookup success in an ACL "recipients" or "senders" condition, or a router
1773 "senders" option, similar to $domain_data and $local_part_data.
1775 6. Moved the writing of debug_print from before to after the "senders" test
1778 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1779 problems for message scanning, either using a data ACL, or using
1780 local_scan() because the Received: header was not generated till after they
1781 were called (in order to set the time as the time of reception completion).
1782 I have revised the way this works. The header is now generated after the
1783 body is received, but before the ACL or local_scan() are called. After they
1784 are run, the timestamp in the header is updated.
1790 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1791 before starting a queue runner without re-exec. This happened only when
1792 deliver_drop_privilege was set or when the Exim user was set to root. The
1793 effect of the bug was that timeouts during subsequent deliveries caused
1794 crashes instead of being properly handled. The handler is now left at its
1795 default (and expected) setting.
1797 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1798 message, again when deliver_drop_privilege is set or Exim is run as root.
1799 The bug described in (1) was not present in this case, but the tidying up
1800 of the other signals was missing. I have made the two cases consistent.
1802 3. The ignore_target_hosts setting on a manualroute router was being ignored
1803 for hosts that were looked up using the /MX notation.
1805 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1808 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1809 operated on the sender address. After changing the $sender_address to <>
1810 for the sender address verify, Exim was re-instated it as the original
1811 (before rewriting) address, but remembering that it had rewritten it, so it
1812 wasn't rewriting it again. This bug also had the effect of breaking the
1813 sender address verification caching when the sender address was rewritten.
1815 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1816 This has been changed so that if the ip literal address matches
1817 ignore_target_hosts, the router declines.
1819 7. Added expansion conditions match_domain, match_address, and match_local_
1820 part (NOT match_host).
1822 8. The placeholder for the Received: header didn't have a length field set.
1824 9. Added code to Exim itself and to exim_lock to test for a specific race
1825 condition that could lead to file corruption when using MBX delivery. The
1826 issue is with the lockfile that is created in /tmp. If this file is removed
1827 after a process has opened it but before that process has acquired a lock,
1828 there is the potential for a second process to recreate the file and also
1829 acquire a lock. This could lead to two Exim processes writing to the file
1830 at the same time. The added code performs the same test as UW imapd; it
1831 checks after acquiring the lock that its file descriptor still refers to
1832 the same named file.
1834 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1835 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1838 11. Added the smtp_active_hostname option. If used, this will typically be made
1839 to depend on the incoming interface address. Because $interface_address is
1840 not set up until the daemon has forked a reception process, error responses
1841 that can happen earlier (such as "too many connections") no longer contain
1844 12. If an expansion in a condition on a "warn" statement fails because a lookup
1845 defers, the "warn" statement is abandoned, and the next ACL statement is
1846 processed. Previously this caused the whole ACL to be aborted.
1848 13. Added the iplsearch lookup type.
1850 14. Added ident_timeout as a log selector.
1852 15. Added tls_certificate_verified as a log selector.
1854 16. Added a global option tls_require_ciphers (compare the smtp transport
1855 option of the same name). This controls incoming TLS connections.
1857 17. I finally figured out how to make tls_require_ciphers do a similar thing
1858 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1859 before starting the TLS session.
1861 18. Tabs are now shown as \t in -bP output.
1863 19. If the log selector return_path_on_delivery was set, Exim crashed when
1864 bouncing a message because it had too many Received: header lines.
1866 20. If two routers both had headers_remove settings, and the first one included
1867 a superfluous trailing colon, the final name in the first list and the
1868 first name in the second list were incorrectly joined into one item (with a
1869 colon in the middle).
1875 1. Added -C and -D options to the exinext utility, mainly to make it easier
1876 to include in the automated testing, but these could be helpful when
1877 multiple configurations are in use.
1879 2. The exinext utility was not formatting the output nicely when there was
1880 an alternate port involved in the retry record key, nor when there was a
1881 message id as well (for retries that were specific to a specific message
1882 and a specific host). It was also confused by IPv6 addresses, because of
1883 the additional colons they contain. I have fixed the IPv4 problem, and
1884 patched it up to do a reasonable job for IPv6.
1886 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1887 delivery, the log line now contains "pipelined" if PIPELINING was used.
1889 4. An SMTP transport process used to panic and die if the bind() call to set
1890 an explicit outgoing interface failed. This has been changed; it is now
1891 treated in the same way as a connect() failure.
1893 5. A reference to $sender_host_name in the part of a conditional expansion
1894 that was being skipped was still causing a DNS lookup. This no longer
1897 6. The def: expansion condition was not recognizing references to header lines
1898 that used bh_ and bheader_.
1900 7. Added the _cache feature to named lists.
1902 8. The code for checking quota_filecount in the appendfile transport was
1903 allowing one more file than it should have been.
1905 9. For compatibility with Sendmail, the command line option
1913 and sets the incoming protocol and host name (for trusted callers). The
1914 host name and its colon can be omitted when only the protocol is to be set.
1915 Note the Exim already has two private options, -pd and -ps, that refer to
1916 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1917 "s", but I don't think that's a major issue.
1919 10. A number of refactoring changes to the code, none of which should affect
1922 (a) The number of logging options was getting close to filling up the
1923 32-bit word that was used as a bit map. I have split them into two classes:
1924 those that are passed in the argument to log_write(), and those that are
1925 only ever tested independently outside of that function. These are now in
1926 separate 32-bit words, so there is plenty of room for expansion again.
1927 There is no change in the user interface or the logging behaviour.
1929 (b) When building, for example, log lines, the code previously used a
1930 macro that called string_cat() twice, in order to add two strings. This is
1931 not really sufficiently general. Furthermore, there was one instance where
1932 it was actually wrong because one of the argument was used twice, and in
1933 one call a function was used. (As it happened, calling the function twice
1934 did not affect the overall behaviour.) The macro has been replaced by a
1935 function that can join an arbitrary number of extra strings onto a growing
1938 (c) The code for expansion conditions now uses a table and a binary chop
1939 instead of a serial search (which was left over from when there were very
1940 few conditions). Also, it now recognizes conditions like "pam" even when
1941 the relevant support is not compiled in: a suitably worded error message is
1942 given if an attempt is made to use such a condition.
1944 11. Added ${time_interval:xxxxx}.
1946 12. A bug was causing one of the ddress fields not to be passed back correctly
1947 from remote delivery subprocesses. The field in question was not being
1948 subsequently used, so this caused to problems in practice.
1950 13. Added new log selectors queue_time and deliver_time.
1952 14. Might have fixed a bug in maildirsizefile handling that threw up
1953 "unexpected character" debug warnings, and recalculated the data
1954 unnecessarily. In any case, I expanded the warning message to give more
1957 15. Added the message "Restricted characters in address" to the statements in
1958 the default ACL that block characters like @ and % in local parts.
1960 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1961 Three changes have been made:
1963 (a) There was a serious bug; a negative response to MAIL caused the whole
1964 recipient domain to be cached as invalid, thereby blocking all messages
1965 to all local parts at the same domain, from all senders. This bug has
1966 been fixed. The domain is no longer cached after a negative response to
1967 MAIL if the sender used is not empty.
1969 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1972 (c) A new callout option, "use_sender" has been added for people who want
1973 the modified behaviour.
1979 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1982 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1983 indeed breaks things for older releases.
1985 3. Added additional logging to the case where there is a problem reading data
1986 from a filter that is running in a subprocess using a pipe, in order to
1987 try to track down a specific problem.
1989 4. Testing facility fudge: when running in the test harness and attempting
1990 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1991 getting "No route to host". Convert this to a timeout.
1993 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1996 6. Some OS don't have socklen_t but use size_t instead. This affects the
1997 fifth argument of getsockopt() amongst other things. This is now
1998 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1999 can be set for individual OS. I have set it for SunOS5, OSF1, and
2000 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2001 some earlier ones do not.
2003 7. Change 4.30/15 was not doing the test caselessly.
2005 8. The standard form for an IPv6 address literal was being rejected by address
2006 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2007 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2008 this, as well as the form without the "IPv6" on the front (but only when
2009 address literals are enabled, of course).
2011 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2013 10. Exim crashed if a message with an empty sender address specified by -f
2014 encountered a router with an errors_to setting. This could be provoked only
2015 by a command such as
2019 where an empty string was supplied; "<>" did not hit this bug.
2021 11. Installed PCRE release 4.5.
2023 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2024 remained set. It is now erased.
2026 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2027 times from message ids (which are base 36 rather than the normal 62).
2029 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2030 were being counted as actual protocol errors, and logged if the log
2031 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2032 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2033 and DATA following a set of rejected RCPTs do not count as protocol errors.
2034 In other words, Exim assumes they were pipelined, though this may not
2035 actually be the case. Of course, in all cases the client gets an
2036 appropriate error code.
2038 15. If a lookup fails in an ACL condition, a message about the failure may
2039 be available; it is used if testing the ACL cannot continue, because most
2040 such messages specify what the cause of the deferral is. However, some
2041 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2042 that caused an old message to be retained and used if a later statement
2043 caused a defer, replacing the real cause of the deferral.
2045 16. If an IP address had so many PTR records that the DNS lookup buffer
2046 was not large enough to hold them, Exim could crash while trying to process
2047 the truncated data. It now detects and logs this case.
2049 17. Further to 4.21/58, another change has been made: if (and only if) the
2050 first line of a message (the first header line) ends with CRLF, a bare LF
2051 in a subsequent header line has a space inserted after it, so as not to
2052 terminate the header.
2054 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2055 unnecessarily, used atoi() instead of strtol(), and didn't check the
2056 termination when getting file sizes from file names by regex.
2058 19. Completely re-implemented the support for maildirsize files, in the light
2059 of a number of problems with the previous contributed implementation
2060 (4.30/29). In particular:
2062 . If the quota is zero, the maildirsize file is maintained, but no quota is
2065 . If the maildir directory does not exist, it is created before any attempt
2066 to write a maildirsize file.
2068 . The quota value in the file is just a cache; if the quota is changed in
2069 the transport, the new value overrides.
2071 . A regular expression is available for excluding directories from the
2074 20. The autoreply transport checks the characters in options that define the
2075 message's headers; it allows continued headers, but it was checking with
2076 isspace() after an embedded newline instead of explicitly looking for a
2079 21. If all the "regular" hosts to which an address was routed had passed their
2080 expiry times, and had not reached their retry times, the address was
2081 bounced, even if fallback hosts were defined. Now Exim should go on to try
2084 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2085 equivalent code in the SMTP transport. Some hosts send humungous responses
2086 to HELO/EHLO, more than 1024 it seems.
2088 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2089 gives compiler warnings in some environments. I've now done it "properly",
2092 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2093 (because of problems with the built-in one) was declared to return uschar *
2094 instead of char *, causing compiler failure.
2096 25. Fixed a file descriptor leak when processing alias/forward files.
2098 26. Fixed a minor format string issue in dbfn.c.
2100 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2102 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2103 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2106 29. When a sender address is verified, it is cached, to save repeating the test
2107 when there is more than one recipient in a message. However, when the
2108 verification involves a callout, it is possible for different callout
2109 options to be set for different recipients. It is too complicated to keep
2110 track of this in the cache, so now Exim always runs a verification when a
2111 callout is required, relying on the callout cache for the optimization.
2112 The overhead is duplication of the address routing, but this should not be
2115 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2116 to be verified with callout=postmaster, and the main callout worked but the
2117 postmaster check failed, the verification correctly failed. However, if a
2118 subsequent RCPT command asked for sender verification *without* the
2119 postmaster check, incorrect caching caused this verification also to fail,
2122 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2123 it was not caching the DNS options (qualify_single, search_parents) that
2124 were used when the lookup failed. A subsequent lookup with different
2125 options therefore always gave the same answer, though there were cases
2126 where it should not have. (Example: a "domains = !$mx_any" option on a
2127 dnslookup router: the "domains" option is always processed without any
2128 widening, but the router might have qualify_single set.) Now Exim uses the
2129 cached value only when the same options are set.
2131 32. Added John Jetmore's "exipick" utility to the distribution.
2133 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2134 than a timeout (e.g. a certificate is required, and is not provided), an
2135 Exim server now closes the connection immediately. Previously it waited for
2136 the client to close - but if the client is SSL, it seems that they each
2137 wait for each other, leading to a delay before one of them times out.
2139 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2140 maintained 0.8.x compatibility because I don't think many are using it, and
2141 it is clearly obsolete.
2143 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2146 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2147 client certificate was expired. A simple patch fixes this, though I don't
2148 understand the full logic of why the verify callback is called multiple
2151 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2152 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2153 which causes problems with some clients (such as the Certicom SSL Plus
2154 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2155 disables the coutermeasure allowing Eudora to connect."
2157 38. Exim was not checking that a write() to a log file succeeded. This could
2158 lead to Bad Things if a log got too big, in particular if it hit a file
2159 size limit. Exim now panics and dies if it cannot write to a log file, just
2160 as it does if it cannot open a log file.
2162 39. Modified OS/Makefile-Linux so that it now contains
2164 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2166 The two -D definitions ensure that Exim is compiled with large file
2167 support, which makes it possible to handle log files that are bigger than
2170 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2171 instance) a domain was checked against a named list that involved a lookup,
2172 causing $domain_data to be set, then another domain was checked against the
2173 same list, then the first domain was re-checked, the value of $domain_data
2174 after the final check could be wrong. In particular, if the second check
2175 failed, it could be set empty. This bug probably also applied to
2178 41. The strip_trailing_dot option was not being applied to the address given
2179 with the -f command-line option.
2181 42. The code for reading a message's header from the spool was incrementing
2182 $received_count, but never initializing it. This meant that the value was
2183 incorrect (doubled) while delivering a message in the same process in which
2184 it was received. In the most common configuration of Exim, this never
2185 happens - a fresh exec is done - but it can happen when
2186 deliver_drop_privilege is set.
2188 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2189 it now includes up to 150 characters of the unexpected data in the log
2192 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2193 and building data strings. The size of both of these buffers was 10 000
2194 bytes - far larger than anybody would *ever* want, thought I. Needless to
2195 say, somebody hit the limit. I have increased the maximum line length to
2196 20 000 and the maximum data length of concatenated lines to 100 000. I have
2197 also fixed two bugs, because there was no checking on these buffers. Tsk,
2198 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2199 buffer is too small.
2201 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2202 lsearch lookups. Now it does.
2204 46. When parsing a route_list item in a manualroute router, a fixed-length
2205 buffer was used for the list of hosts. I made this 1024 bytes long,
2206 thinking that nobody would ever have a list of hosts that long. Wrong.
2207 Somebody had a whole pile of complicated expansion conditions, and the
2208 string was silently truncated, leading to an expansion error. It turns out
2209 that it is easier to change to an unlimited length (owing to other changes
2210 that have happened since this code was originally written) than to build
2211 structure for giving a limitation error. The length of the item that
2212 expands into the list of hosts is now unlimited.
2214 47. The lsearch lookup could not handle data where the length of text line was
2215 more than 4095 characters. Such lines were truncated, leading to shortened
2216 data being returned. It should now handle lines of any length.
2218 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2219 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2222 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2223 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2224 no longer happens in comment lines. A list of replacements is now placed
2225 at the head of all of the source files, except those whose only change is
2226 to replace PERL_COMMAND in the very first #! line.
2228 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2229 messages on the queue, with a bottom-up merge sort, using code contributed
2230 by Michael Haardt. This should make operations like -bp somewhat faster on
2231 large queues. It won't affect queue runners, except when queue_run_in_order
2234 51. Installed eximstats 1.31 in the distribution.
2236 52. Added support for SRV lookups to the dnslookup router.
2238 53. If an ACL referred to $message_body or $message_body_end, the value was not
2239 reset for any messages that followed in the same SMTP session.
2241 54. The store-handling optimization for building very long strings was not
2242 differentiating between the different store pools. I don't think this
2243 actually made any difference in practice, but I've tidied it.
2245 55. While running the routers to verify a sender address, $sender_address
2246 was still set to the sender address. This is wrong, because when routing to
2247 send a bounce to the sender, it would be empty. Therefore, I have changed
2248 it so that, while verifying a sender address, $sender_address is set to <>.
2249 (There is no change to what happens when verifying a recipient address.)
2251 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2252 target A or AAAA records (if not already returned) without resetting the
2253 qualify_single or search_parents options of the DNS resolver. These are
2254 inappropriate in this case because the targets of MX and SRV records must
2255 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2256 target that, when qualified, matched something in the local domain. These
2257 two options are now turned off when doing these lookups.
2259 57. It seems that at least some releases of Reiserfs (which does not have the
2260 concept of a fixed number of inodes) returns zero and not -1 for the
2261 number of available inodes. This interacted badly with check_spool_inodes,
2262 which assumed that -1 was the "no such thing" setting. What I have done is
2263 to check that the total number of inodes is greater than zero before doing
2264 the test of how many are available.
2266 58. When a "warn" ACL statement has a log_message modifier, the message is
2267 remembered, and not repeated. This is to avoid a lot of repetition when a
2268 message has many recipients that cause the same warning to be written.
2269 Howewer, Exim was preserving the list of already written lines for an
2270 entire SMTP session, which doesn't seem right. The memory is now reset if a
2271 new message is started.
2273 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2274 debugging output unless log_rewrite was also set.
2276 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2277 of (int)(handle) when we know that handle contains (void *)(-1).
2279 61. The Exim daemon panic-logs an error return when it closes the incoming
2280 connection. However "connection reset by peer" seems to be common, and
2281 isn't really an error worthy of noting specially, so that particular error
2284 62. When Exim is trying to find all the local interfaces, it used to panic and
2285 die if the ioctl to get the interface flags failed. However, it seems that
2286 on at least one OS (Solaris 9) it is possible to have an interface that is
2287 included in the list of interfaces, but for which you get a failure error
2288 for this call. This happens when the interface is not "plumbed" into a
2289 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2290 failure of the "get flags" call assumes that the interface is down.
2292 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2293 makes life easier for people who are doing arithmetic on fields extracted
2294 from dates, where you often get leading zeros that should not be
2295 interpreted as octal.
2297 64. Added qualify_domain to the redirect router, to override the global
2300 65. If a pathologically long header line contained very many addresses (the
2301 report of this problem mentioned 10 000) and each of them was rewritten,
2302 Exim could use up a very large amount of memory. (It kept on making new
2303 copies of the header line as it rewrote, and never released the old ones.)
2304 At the expense of a bit more processing, the header rewriting function has
2305 been changed so that it no longer eats memory in this way.
2307 66. The generation of the Received: header has been moved from the time that a
2308 message starts to be received, to the time that it finishes. The timestamp
2309 in the Received: header should now be very close to that of the <= log
2310 line. There are two side-effects of this change:
2312 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2313 logged header lines no longer include the local Received: line, because
2314 it has not yet been created. The same applies to a copy of the message
2315 that is returned to a non-SMTP sender when a message is rejected.
2317 (b) When a filter file is tested using -bf, no additional Received: header
2318 is added to the test message. After some thought, I decided that this
2321 This change does not affect the value of $received_for. It is still set
2322 after address rewriting, but before local_scan() is called.
2324 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2326 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2327 gave an unhelpful panic error message, and a defer error. I have managed to
2328 change this behaviour so that it now rejects any supplied certificate,
2329 which seems right, as the list of acceptable certificates is empty.
2331 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2332 gave an unhelpful defer error. I have not managed to make this reject any
2333 supplied certificates, but the error message it gives is "no certificate
2334 supplied", which is not helpful.
2336 70. exigrep's output now also includes lines that are not associated with any
2337 message, but which match the given pattern. Implemented by a patch from
2338 Martin Sluka, which also tidied up the Perl a bit.
2340 71. Recipient callout verification, like sender verification, was using <> in
2341 the MAIL FROM command. This isn't really the right thing, since the actual
2342 sender may affect whether the remote host accepts the recipient or not. I
2343 have changed it to use the actual sender in the callout; this means that
2344 the cache record is now keyed on a recipient/sender pair, not just the
2345 recipient address. There doesn't seem to be a real danger of callout loops,
2346 since a callout by the remote host to check the sender would use <>.
2347 [SEE ABOVE: changed after hitting problems.]
2349 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2350 temporary errors. However, in the case of such a code being given after
2351 the end of a data transmission (i.e. after ".") Exim was failing to write
2352 a retry record for the message. (Yes, there was some broken host that was
2353 actually sending 8xx at this point.)
2355 73. An unknown lookup type in a host list could cause Exim to panic-die when
2356 the list was checked. (An example that provoked this was putting <; in the
2357 middle of a list instead of at the start.) If this happened during a DATA
2358 ACL check, a -D file could be left lying around. This kind of configuration
2359 error no longer causes Exim to die; instead it causes a defer errror. The
2360 incident is still logged to the main and panic logs.
2362 74. Buglet left over from Exim 3 conversion. The message "too many messages
2363 in one connection" was written to the rejectlog but not the mainlog, except
2364 when address rewriting (yes!) was being logged.
2366 75. Added write_rejectlog option.
2368 76. When a system filter was run not as root (that is, when system_filter_user
2369 was set), the values of the $n variables were not being returned to the
2370 main process; thus, they were not subsequently available in the $sn
2373 77. Added +return_path_on_delivery log selector.
2375 78. A connection timeout was being treated differently from recipients deferred
2376 when testing hosts_max_try with a message that was older than the host's
2377 retry timeout. (The host should not be counted, thus allowing all hosts to
2378 be tried at least once before bouncing.) This may have been the cause of an
2379 occasionally reported bug whereby a message would remain on the queue
2380 longer than the retry timeout, but would be bounced if a delivery was
2381 forced. I say "may" because I never totally pinned down the problem;
2382 setting up timeout/retry tests is difficult. See also the next item.
2384 79. The ultimate address timeout was not being applied to errors that involved
2385 a combination of host plus message (for example, a timeout on a MAIL
2386 command). When an address resolved to a number of possible hosts, and they
2387 were not all tried for each delivery (e.g. because of hosts_max_try), a
2388 message could remain on the queue longer than the retry timeout.
2390 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2393 81. Fixed an obscure SMTP outgoing bug which required at least the following
2394 conditions: (a) there was another message waiting for the same server;
2395 (b) the server returned 5xx to all RCPT commands in the first message so
2396 that the message was not completed; (c) the server dropped the connection
2397 or gave a negative response to the RSET that Exim sends to abort the
2398 transaction. The observed case was a dropped connection after DATA that had
2399 been sent in pipelining mode. That is, the server had advertised PIPELINING
2400 but was not implementing it correctly. The effect of the bug was incorrect
2401 behaviour, such as trying another host, and this could lead to a crash.
2407 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2408 and daemon.c were passed as pointers to ints; they should have been
2409 pointers to socklen_t variables (which are typically unsigned ints).
2411 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2414 3. Fixed a really odd bug that affected only the testing scheme; patching a
2415 certain fixed string in the binary changed the value of another string that
2416 happened to be identical to the end of the original first string.
2418 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2419 name", it returns that address as the IP address. On some operating
2420 systems (e.g. Solaris), it also passes back the IP address string as the
2421 "host name". However, on others (e.g. Linux), it passes back an empty
2422 string. Exim wasn't checking for this, and was changing the host name to an
2423 empty string, assuming it had been canonicized.
2425 5. Although rare, it is permitted to have more than one PTR record for a given
2426 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2427 all the names associated with an address, because they do in Solaris.
2428 However, it seems that they do not in Linux for data that comes from the
2429 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2430 I found this out when I moved to a new Linux workstation and tried to run
2431 the Exim test suite.
2433 To get round this problem I have changed the code so that it now does its
2434 own call to the DNS to look up PTR records when searching for a host name.
2435 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2436 addresses that are only in /etc/hosts are still found.
2438 This behaviour is, however, controlled by an option called host_lookup_
2439 order, which defaults to "bydns:byaddr". If people want to use the other
2440 order, or indeed, just use one or the other means of lookup, they can
2441 specify it in this variable.
2443 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2444 some operating systems, this comes back from gethostbyaddr() as an empty
2445 string, and this is what Exim used to test for. However, it seems that in
2446 other systems, "." is yielded. Exim now tests for this case too.
2448 7. The values of check_spool_space and check_log_space are now held internally
2449 as a number of kilobytes instead of an absolute number of bytes. If a
2450 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2451 kilobyte. This means that much larger values can be stored.
2453 8. Exim monitor: an attempt to get the action menu when not actually pointing
2454 at a message produces an empty menu entitled "No message selected". This
2455 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2456 no entries in it ("Shell widget menu has zero width and/or height"). So I
2457 have added a single, blank menu entry in this case.
2459 9. Added ${quote_local_part.
2461 10. MIME decoding is now applied to the contents of Subject: header lines when
2464 11. Now that a reference to $sender_host_address automatically causes a reverse
2465 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2466 host lookup before query-style lookups in lists that might use this
2467 variable. This has therefore been abolished, and the "net-" prefix is no
2468 longer necessary for query-style lookups.
2470 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2471 have been a typo for LFLAGS, so it has been changed.
2473 13. The install script calls Exim with "-C /dev/null" in order to find the
2474 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2475 to be output. Howeve, since Exim outputs its version number before the
2476 error, it didn't break the script. It just looked ugly. I fixed this by
2477 always allowing "-C /dev/null" if the caller is root.
2479 14. Ignore overlarge ACL variable number when reading spool file - insurance
2480 against a later release with more variables having written the file.
2482 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2483 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2484 this, as well as the form without the "IPv6" on the front.
2486 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2487 OS/Makefile-Darwin file.
2489 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2490 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2492 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2493 variable. However, some OS use ints for inodes. I've added cast to long int
2494 to get rid of the compiler warning.
2496 19. I had forgotten to lock out "/../" in configuration file names when
2497 ALT_CONFIG_PREFIX was set.
2499 20. Routers used for verification do not need to specify transports. However,
2500 if such a router generated a host list, and callout was configured, Exim
2501 crashed, because it could not find a port number from the (non-existent)
2502 transport. It now assumes port 25 in this circumstance.
2504 21. Added the -t option to exigrep.
2506 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2507 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2508 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2509 them is set, LOOKUP_LSEARCH is forced.
2511 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2513 24. Added sender and host information to the "rejected by local_scan()" log
2514 line; previously there was no indication of these.
2516 25. Added .include_if_exists.
2518 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2519 Linux. It turns out that not all file systems support this. Apparently some
2520 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2521 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2522 not supported on the file descriptor, is now ignored when Exim is trying to
2523 sync a directory. This applies only to Linux.
2525 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2527 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2528 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2531 29. Added support for maildirsize files from supplied patch (modified a bit).
2533 30. The use of :fail: followed by an empty string could lead Exim to respond to
2534 sender verification failures with (e.g.):
2536 550 Verification failed for <xxx>
2537 550 Sender verify failed
2539 where the first response line was missing the '-' that indicates it is not
2540 the final line of the response.
2542 31. The loop for finding the name of the user that called Exim had a hardwired
2543 limit of 10; it now uses the value of finduser_retries, which is used for
2544 all other user lookups.
2546 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2549 33. Exim was neglecting to zero errno before one call of strtol() when
2550 expanding a string and expecting an integer value. On some systems this
2551 resulted in spurious "integer overflow" errors. Also, it was casting the
2552 result into an int without checking.
2554 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2555 did not work. The code looks as if it has *never* worked, though it appears
2556 to have been documented since at least releast 1.62. I have made it work.
2558 35. The "timeout_DNS" error in retry rules, also documented since at least
2559 1.62, also never worked. As it isn't clear exactly what this means, and
2560 clearly it isn't a major issue, I have abolished the feature by treating it
2561 as "timeout", and writing a warning to the main and panic logs.
2563 36. The display of retry rules for -brt wasn't always showing the error code
2566 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2567 timeout_connect_A, timeout_connect_MX.
2569 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2570 to the empty sender.
2572 39. The daemon was not analysing the content of -oX till after it had closed
2573 stderr and disconnected from the controlling terminal. This meant that any
2574 syntax errors were only noted on the panic log, and the return code from
2575 the command was 0. By re-arranging the code a little, I've made the
2576 decoding happen first, so such errors now appear on stderr, and the return
2577 code is 1. However, the actual setting up of the sockets still happens in
2578 the disconnected process, so errors there are still only recorded on the
2581 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2582 connections (as happens on some IP stacks) was logged at start up time as
2583 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2584 it from "IPv6 and IPv4", which means that two separate sockets are being
2587 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2588 says whether AF_INET or AF_INET6 was passed as an argument.
2590 42. Exiwhat output was messed up when time zones were included in log
2593 43. Exiwhat now gives more information about the daemon's listening ports,
2594 and whether -tls-on-connect was used.
2596 44. The "port" option of the smtp transport is now expanded.
2598 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2599 silently ignored. Now an error message is written to the main and panic
2602 46. There's a new ACL modifier called "logwrite" which writes to a log file
2603 as soon as it is encountered.
2605 47. Added $local_user_uid and $local_user_gid at routing time.
2607 48. Exim crashed when trying to verify a sender address that was being
2610 49. Exim was recognizing only a space character after ".include". It now also
2611 recognizes a tab character.
2613 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2614 extracting the relevant information from the specification. The man page no
2615 longer contains scrambled data for the -d option, and I've added a section
2616 at the front about calling Exim under different names.
2618 51. Added "extra_headers" argument to the "mail" command in filter files.
2620 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2623 53. Installed eximstats 1.29.
2625 54. Added transport_filter_timeout as a generic transport option.
2627 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2628 Cc: header lines. This was required by RFC 822, but it not required by RFC
2631 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2632 incoming messages that did not have them. Now it does so only if the
2633 message originates locally, that is, if there is no associated remote host
2634 address. When Resent- header lines are present, this applies to the Resent-
2635 lines rather than the non-Resent- lines.
2637 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2638 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2640 58. Messages for configuration errors now include the name of the main
2641 configuration file - useful now that there may be more than one file in a
2642 list (.included file names were always shown).
2644 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2645 for those rare installations that do not start the daemon as root or run it
2646 setuid root. I've cut out the call to initgroups() if the daemon is not
2649 60. The Exim user and group can now be bound into the binary as text strings
2650 that are looked up at the start of Exim's processing.
2652 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2654 62. Added $mailstore_basename variable.
2656 63. Installed patch to sieve.c from Michael Haardt.
2658 64. When Exim failed to open the panic log after failing to open the main log,
2659 the original message it was trying to log was written to stderr and debug
2660 output, but if they were not available (the usual case in production), it
2661 was lost. Now it is written to syslog before the two lines that record the
2662 failures to open the logs.
2664 65. Users' Exim filters run in subprocesses under the user's uid. It is
2665 possible for a "deliver" command or an alias in a "personal" command to
2666 provoke an address rewrite. If logging of address rewriting is configured,
2667 this fails because the process is not running as root or exim. There may be
2668 a better way of dealing with this, but for the moment (because 4.30 needs
2669 to be released), I have disabled address rewrite logging when running a
2670 filter in a non-root, non-exim process.
2676 1. The buildconfig auxiliary program wasn't quoting the value set for
2677 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2678 not defined. This bug was masked in 4.22 by the effect that was fixed in
2681 2. Some messages that were rejected after a message id was allocated were
2682 shown as "incomplete" by exigrep. It no longer does this for messages that
2683 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2685 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2686 have allow_domain_literals set, the ID did not get logged in the <= line.
2687 Domain literals are now always recognized in Message-ID: header lines.
2689 4. The first argument for a ${extract expansion item is the key name or field
2690 number. Leading and trailing spaces in this item were not being ignored,
2691 causing some misleading effects.
2693 5. When deliver_drop_privilege was set, single queue runner processes started
2694 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2695 same command in the process it spins off) were not dropping privilege.
2697 6. When the daemon running as "exim" started a queue runner, it always
2698 re-executed Exim in the spun-off process. This is a waste of effort when
2699 deliver_drop_privilege is set. The new process now just calls the
2700 queue-runner function directly.
2706 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2709 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2710 ignored. Though the use of -oP was forcing the writing of a pid file, it
2711 was always written to the default place.
2713 3. If the message "no IP address found for host xxxx" is generated during
2714 incoming verification, it is now followed by identification of the incoming
2715 connection (so you can more easily find what provoked it).
2717 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2719 5. Added some features to "harden" Exim a bit more against certain attacks:
2721 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2722 be put in Local/Makefile. This is like the never_users runtime option,
2723 but it cannot be overridden. The default setting is "root".
2725 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2726 prefix string with which any file named in a -C command line option
2729 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2730 is retained for -C and -D only if the caller of Exim is root. Without
2731 it, the exim user may also use -C and -D and retain privilege.
2733 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2734 command line option is disabled.
2736 6. Macro names set by the -D option must start with an upper case letter, just
2737 like macro names defined in the configuration file.
2739 7. Added "dereference=" facility to LDAP.
2741 8. Two instances of the typo "uknown" in the source files are fixed.
2743 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2744 the Configure-Makefile script screwed up while processing it.
2746 10. Incorporated PCRE 4.4.
2748 11. The SMTP synchronization check was not operating right at the start of an
2749 SMTP session. For example, it could not catch a HELO sent before the client
2750 waited for the greeting. There is now a check for outstanding input at the
2751 point when the greeting is written. Because of the duplex, asynchronous
2752 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2753 way, but not yet received, when the check is performed.
2755 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2756 on TCP/IP sockets, because this apparently causes some broken clients to
2759 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2760 unchanged) from the Cygwin maintainer.
2762 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2763 was supported instead of testing for mbx. Effectively a typo.
2765 15. The spa authenticator server code was not checking that the input it
2766 received was valid base64.
2768 16. The debug output line for the "set" modifier in ACLs was not showing the
2769 name of the variable that was being set.
2771 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2773 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2774 lookup on its own. Something else was needed to trigger the lookup. For
2775 example, a match in host_lookup or the need for a host name in a host list.
2776 Now, if $sender_host_name is referenced and the host name has not yet been
2777 looked up, a lookup is performed. If the lookup fails, the variable remains
2778 empty, and $host_lookup_failed is set to "1".
2780 19. Added "eqi" as a case-independent comparison operator.
2782 20. The saslauthd authentication condition could segfault if neither service
2783 nor realm was specified.
2785 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2786 error message that was logged was misleading, and incoming SMTP
2787 connections were dropped. The message is now more accurate, and temporary
2788 errors are given to SMTP connections.
2790 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2791 (except RSET) with a 421 error, until QUIT is received. However, it was
2792 failing to send a response to QUIT.
2794 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2795 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2796 ACL is now run after the helo_try_verify_hosts code.
2798 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2799 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2800 case-independent, so other case variants are also recognized). Apparently
2801 some systems use these upper case variants.
2803 25. If more than two messages were waiting for the same host, and a transport
2804 filter was specified for the transport, Exim sent two messages over the
2805 same TCP/IP connection, and then failed with "socket operation on non-
2806 socket" when it tried to send the third.
2808 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2810 27. The extern definition of crypt16() in expand.c was not being excluded when
2811 the OS had its own crypt16() function.
2813 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2814 as a preferred synonym for return_size_limit, both as an option and as an
2817 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2819 30. Changed the default configuration ACL to relax the local part checking rule
2820 for addresses that are not in any local domains. For these addresses,
2821 slashes and pipe symbols are allowed within local parts, but the sequence
2822 /../ is explicitly forbidden.
2824 31. SPA server authentication was not clearing the challenge buffer before
2827 32. log_message in a "warn" ACL statement was writing to the reject log as
2828 well as to the main log, which contradicts the documentation and doesn't
2829 seem right (because no rejection is happening). So I have stopped it.
2831 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2832 However, I am unable to do any testing of this.
2834 34. Fixed an infelicity in the appendfile transport. When checking directories
2835 for a mailbox, to see if any needed to be created, it was accidentally
2836 using path names with one or more superfluous leading slashes; tracing
2837 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2839 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2840 contents are added to the log line that is written for every discarded
2841 recipient. (Previously a log_message setting was ignored.)
2843 36. The ${quote: operator now quotes the string if it is empty.
2845 37. The install script runs exim in order to find its version number. If for
2846 some reason other than non-existence or emptiness, which it checks, it
2847 could not run './exim', it was installing it with an empty version number,
2848 i.e. as "exim-". This error state is now caught, and the installation is
2851 38. An argument was missing from the function that creates an error message
2852 when Exim fails to connect to the socket for saslauthd authentication.
2853 This could cause Exim to crash, or give a corrupted message.
2855 39. Added isip, isip4, and isip6 to ${if conditions.
2857 40. The ACL variables $acl_xx are now saved with the message, and can be
2858 accessed later in routers, transports, and filters.
2860 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2861 strings in the file are not string-expanded.
2863 42. If a MAIL command specified a SIZE value that was too large to fit into an
2864 int variable, the check against message_size_limit failed. Such values are
2865 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2866 day this will have to be increased, but I don't think I want to be around
2867 when emails are that large.
2874 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2875 iconv() is not standard in FreeBSD.
2877 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2878 IPv6 enabled. The observed symptom was a segmentation fault on return from
2879 the function os_common_find_running_interfaces() in src/os.c.
2881 3. In the check_special_case() function in daemon.c I had used "errno" as an
2882 argument name, which causes warnings on some systems. This was basically a
2883 typo, since it was named "eno" in the comments!
2885 4. The code that waits for the clock to tick (at a resolution of some fraction
2886 of a second) so as to ensure message-id uniqueness was always waiting for
2887 at least one whole tick, when it could have waited for less. [This is
2888 almost certainly not relevant at current processor speeds, where it is
2889 unlikely to ever wait at all. But we try to future-proof.]
2891 5. The function that sleeps for a time interval that includes fractions of a
2892 second contained a race. It did not block SIGALRM between setting the
2893 timer, and suspending (a couple of lines later). If the interval was short
2894 and the sigsuspend() was delayed until after it had expired, the suspension
2895 never ended. On busy systems this could lead to processes getting stuck for
2898 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2899 process, before it forks any delivery processes. The open lookup caching
2900 mechanism meant that the open file or database connection was passed into
2901 the delivery process. The problem was that delivery processes always tidy
2902 up cached lookup data. This could cause a problem for the next delivery
2903 process started by the queue runner, because the external queue runner
2904 process does not know about the closure. So the next delivery process
2905 still has data in the lookup cache. In the case of a file lookup, there was
2906 no problem because closing a file descriptor in a subprocess doesn't affect
2907 the parent. However, if the lookup was caching a connection to a database,
2908 the connection was closed, and the second delivery process was likely to
2909 see errors such as "PGSQL: query failed: server closed the connection
2910 unexpectedly". The problem has been fixed by closing all cached lookups
2911 in a queue runner before running a delivery process.
2913 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2914 seem to have the "const" qualifier which it has on other OS. I've
2917 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2918 *of the same type* (client or server) with the same public name that an
2919 error should be diagnosed.
2921 9. When Exim looked up a host name for an IP address, but failed to find the
2922 original IP address when looking up the host name (a safety check), it
2923 output the message "<ip address> does not match any IP for NULL", which was
2924 confusing, to say the least. The bug was that the host name should have
2925 appeared instead of "NULL".
2927 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2928 user that is built into the binary, and the -C or -D options is used, root
2929 privilege is dropped before the configuration file is read. In addition,
2930 logging is switched to stderr instead of the normal log files. If the
2931 configuration then re-defines the Exim user, the unprivileged environment
2932 is probably not what is expected, so Exim logs a panic warning message (but
2935 However, if deliver_drop_privilege is set, the unprivileged state may well
2936 be exactly what is intended, so the warning has been cut out in that case,
2937 and Exim is allowed to try to write to its normal log files.
2943 1. smtp_return_error_details was not giving details for temporary sender
2944 or receiver verification errors.
2946 2. Diagnose a configuration error if two authenticators have the same public
2949 3. Exim used not to create the message log file for a message until the first
2950 delivery attempt. This could be confusing when incoming messages were held
2951 for policy or load reasons. The message log file is now created at the time
2952 the message is received, and an initial "Received" line is written to it.
2954 4. The automatically generated man page for command line options had a minor
2955 bug that caused no ill effects; however, a more serious problem was that
2956 the procedure for building the man page automatically didn't always
2957 operate. Consequently, release 4.20 contains an out-of-date version. This
2958 shouldn't happen again.
2960 5. When building Exim with embedded Perl support, the script that builds the
2961 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2962 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2964 6. The freeze_tell option was not being used for messages that were frozen on
2965 arrival, either by an ACL or by local_scan().
2967 7. Added the smtp_incomplete_transaction log selector.
2969 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2970 was accepting AUTH without a new EHLO.
2972 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2973 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2976 10. Logging of TCP/IP connections (when configured) now happens in the main
2977 daemon process instead of the child process, so that the TCP/IP connection
2978 count is more accurate (but it can never be perfect).
2980 11. The use of "drop" in a nested ACL was not being handled correctly in the
2981 outer ACL. Now, if condition failure induced by the nested "drop" causes
2982 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2983 or "require"), the connection is dropped.
2985 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2986 that yield "discard" can now be used with an "accept" or a "discard" verb,
2987 but an error is generated for any others (because I can't see a useful way
2988 to define what should happen).
2990 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2991 are now processed in the same way as lines in the Exim configuration file.
2992 In particular, continuation lines are supported.
2994 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2996 15. Added -ti meaning -t -i.
2998 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2999 domains, and warn by logging if others are found.
3001 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3002 returned from GIFCONF when Exim is trying to find the list of interfaces on
3003 a host. The code in os.c has been modified to copy each ifreq to an aligned
3004 structure in all cases.
3006 Also, in some cases, the returned ifreq's were being copied to a 'struct
3007 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3008 means the last couple of bytes of an IPv6 address could be chopped if the
3009 ifreq contained only a normal sockaddr (14 bytes storage).
3011 18. Named domain lists were not supported in the hosts_treat_as_local option.
3012 An entry such as +xxxx was not recognized, and was treated as a literal
3015 19. Ensure that header lines added by a DATA ACL are included in the reject log
3016 if the ACL subsequently rejects the message.
3018 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3019 MD5 (which is deprecated).
3021 21. When testing a filter file using -bf, Exim was writing a message when it
3022 took the sender from a "From " line in the message, but it was not doing so
3023 when it took $return_path from a Return-Path: header line. It now does.
3025 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3026 with a valid header line field name (a series of printing characters
3027 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3029 23. Changed "disc" in the source to "disk" to conform to the documentation and
3030 the book and for uniformity.
3032 24. Ignore Sendmail's -Ooption=value command line item.
3034 25. When execve() failed while trying to run a command in a pipe transport,
3035 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3036 could be confused with a return value of 69 from the command itself. This
3037 has been changed to 127, the value the shell returns if it is asked to run
3038 a non-existent command. The wording for the related log line suggests a
3039 non-existent command as the problem.
3041 26. If received_header_text expands to an empty string, do not add a Received:
3042 header line to the message. (Well, it adds a token one on the spool, but
3043 marks it "old" so that it doesn't get used or transmitted.)
3045 27. Installed eximstats 1.28 (addition of -nt option).
3047 28. There was no check for failure on the call to getsockname() in the daemon
3048 code. This can fail if there is a shortage of resources on the system, with
3049 ENOMEM, for example. A temporary error is now given on failure.
3051 29. Contrary to the C standard, it seems that in some environments, the
3052 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3053 program. Exim now does this explicitly; it affects the formatting of
3054 timestamps using strftime().
3056 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3057 complaints. I've now initialized all the variables, to avoid this.
3059 32. Header lines added by a system filter were not being "seen" during
3060 transport-time rewrites.
3062 33. The info_callback() function passed to OpenSSL is set up with type void
3063 (*)(SSL *, int, int), as described somewhere. However, when calling the
3064 function (actually a macro) that sets it up, the type void(*)() is
3065 expected. I've put in a cast to prevent warnings from picky compilers.
3067 34. If a DNS black list lookup found a CNAME record, but there were no A
3068 records associated with the domain it pointed at, Exim crashed.
3070 35. If a DNS black list lookup returned more than one A record, Exim ignored
3071 all but the first. It now scans all returned addresses if a particular IP
3072 value is being sought. In this situation, the contents of the
3073 $dnslist_value variable are a list of all the addresses, separated by a
3076 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3077 to accept a host name and all its aliases if the forward lookup for any of
3078 them yielded the IP address of the incoming connection. Now it accepts only
3079 those names whose forward lookup yields the correct IP address. Any other
3080 names are discarded. This closes a loophole whereby a rogue DNS
3081 administrator could create reverse DNS records to break through a
3082 wildcarded host restriction in an ACL.
3084 37. If a user filter or a system filter that ran in a subprocess used any of
3085 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3086 the wrong values were passed to the pipe command ($thisaddress had the
3087 value of $0, $0 had the value of $1, etc). This bug was introduced by
3088 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3090 38. Improved the line breaking for long SMTP error messages from ACLs.
3091 Previously, if there was no break point between 40 and 75 characters, Exim
3092 left the rest of the message alone. Two changes have been made: (a) I've
3093 reduced the minimum length to 35 characters; (b) if it can't find a break
3094 point between 35 and 75 characters, it looks ahead and uses the first one
3095 that it finds. This may give the occasional overlong line, but at least the
3096 remaining text gets split now.
3098 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3099 file descriptors might be low, and that setting 1000 would always raise it.
3100 It turns out that in some environments, the limit is already over 1000 and
3101 that lowering it causes trouble. So now Exim takes care not to decrease it.
3103 40. When delivering a message, the value of $return_path is set to $sender_
3104 address at the start of routing (routers may change the value). By an
3105 oversight, this default was not being set up when an address was tested by
3106 -bt or -bv, which affected the outcome if any router or filter referred to
3109 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3110 log only when "message" or "log_message" are set. However, if one of the
3111 conditions was an address verification, or a call to a nested ACL, the
3112 messages generated by the underlying test were being passed through. This
3113 no longer happens. The underlying message is available in $acl_verify_
3114 message for both "message" and "log_message" expansions, so it can be
3115 passed through if needed.
3117 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3118 new expansion $bh_ to give the encoded byte string without charset
3119 translation. Translation happens only if iconv() is available; HAVE_ICONV
3120 indicates this at build time. HEADERS_CHARSET gives the charset to
3121 translate to; headers_charset can change it in the configuration, and
3122 "headers charset" can change it in an individual filter file.
3124 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3125 that creates RFC 2047 encoded "words" labels them as that charset instead
3126 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3127 expansion operator; (ii) when Exim creates a From: line for a local
3128 message; (iii) when a header line is rewritten to include a "phrase" part.
3130 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3131 buggy, causing it to skip the first lines of messages whose message ID
3132 ended in 'D'. This would not have bitten before Exim release 4.14, because
3133 message IDs were unlikely to end in 'D' before then. The effect was to have
3134 incorrect size information for certain domains.
3136 45. #include "config.h" was missing at the start of the crypt16.c module. This
3137 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3140 46. If there was a timeout during a "random" callout check, Exim treated it as
3141 a failure of the random address, and carried on sending RSET and the real
3142 address. If the delay was just some slowness somewhere, the response to the
3143 original RCPT would be taken as a response to RSET and so on, causing
3144 mayhem of various kinds.
3146 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3147 when I implemented it. It didn't allow for the fact that some option values
3148 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3149 the right test for positive values.
3151 48. Domain names in DNS records are case-independent. Exim always looks them up
3152 in lower case. Some resolvers return domain names in exactly the case they
3153 appear in the zone file, that is, they may contain uppercase letters. Not
3154 all resolvers do this - some return always lower case. Exim was treating a
3155 change of case by a resolver as a change of domain, similar to a widening
3156 of a domain abbreviation. This triggered its re-routing code and so it was
3157 trying to route what was effectively the same domain again. This normally
3158 caused routing to fail (because the router wouldn't handle the domain
3159 twice). Now Exim checks for this case specially, and just changes the
3160 casing of the domain that it ultimately uses when it transmits the message
3163 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3166 50. If a filter generated a file delivery with a non-absolute name (possible if
3167 no home directory exists for the router), the forbid_file option was not
3170 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3171 the existing equality matching.
3173 52. Exim was using ints instead of ino_t variables in some places where it was
3174 dealing with inode numbers.
3176 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3177 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3178 called TMPDIR, and if it finds it is different, it changes its value.
3180 54. The smtp_printf() function is now made available to local_scan() so
3181 additional output lines can be written before returning. There is also an
3182 smtp_fflush() function to enable the detection of a dropped connection.
3183 The variables smtp_input and smtp_batched_input are exported to
3186 55. Changed the default runtime configuration: the message "Unknown user"
3187 has been removed from the ACL, and instead placed on the localuser router,
3188 using the cannot_route_message feature. This means that any verification
3189 failures that generate their own messages won't get overridden. Similarly,
3190 the "Unrouteable address" message that was in the ACL for unverifiable
3191 relay addresses has also been removed.
3193 56. Added hosts_avoid_esmtp to the smtp transport.
3195 57. The exicyclog script was not checking for the esoteric option
3196 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3197 will work only if exicyclog is run under the appropriate euid.
3199 58. Following a discussion on the list, the rules by which Exim recognises line
3200 endings on incoming messages have been changed. The -dropcr and drop_cr
3201 options are now no-ops, retained only for backwards compatibility. The
3202 following line terminators are recognized: LF CRLF CR. However, special
3203 processing applies to CR:
3205 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3206 nor a local message in the state where . is a terminator.
3208 (ii) If a bare CR is encountered in a header line, an extra space is added
3209 after the line terminator so as not to end the header. The reasoning
3210 behind this is that bare CRs in header lines are most likely either
3211 to be mistakes, or people trying to play silly games.
3213 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3214 was being incorrectly given as 18 bytes larger than it should have been.
3215 This is a VOB (very old bug).
3217 60. This may never have affected anything current, but just in case it has:
3218 When the local host is found other than at the start of a list of hosts,
3219 the local host, those with the same MX, and any that follow, are discarded.
3220 When the list in question was part of a longer list of hosts, the following
3221 hosts (not currently being processed) were also being discarded. This no
3222 longer happens. I'm not sure if this situation could ever has previously
3225 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3228 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3229 header. This is something that is recommended in a new Internet Draft, and
3230 is something that is documented as being done by Sendmail. There are two
3231 possible values. For messages generated by the autoreply transport, Exim
3234 Auto-Submitted: auto-replied
3236 whereas for all other generated messages (e.g. bounces) it adds
3238 Auto-Submitted: auto-generated
3240 63. The "personal" condition in filters now includes a test for the
3241 Auto-Submitted: header. If it contains the string "auto-" the message it
3242 not considered personal.
3244 64. Added rcpt_include_affixes as a generic transport option.
3246 65. Added queue_only_override (default true).
3248 66. Added the syslog_duplication option.
3250 67. If what should have been the first header line of a message consisted of
3251 a space followed by a colon, Exim was mis-interpreting it as a header line.
3252 It isn't of course - it is syntactically invalid and should therefore be
3253 treated as the start of the message body. The misbehaviour could have
3254 caused a number of strange effects, including loss of data in subsequent
3255 header lines, and spool format errors.
3257 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3258 client host had authenticated. This control can now be exercised by an ACL
3259 for more flexibility.
3261 69. By default, callouts do not happen when testing with -bh. There is now a
3262 variant, -bhc, which does actually run the callout code, including
3263 consulting and updating the callout cache.
3265 70. Added support for saslauthd authentication, courtesy of Alexander
3268 71. If statvfs() failed on the spool or log directories while checking their
3269 size for availability, Exim confusingly gave the error "space shortage".
3270 Furthermore, in debugging mode it crashed with a floating point exception.
3271 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3272 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3273 serious problem, Exim now writes to the main and panic logs when this
3274 happens, with details of the failure. It then refuses to accept the
3275 incoming message, giving the message "spool directory problem" or "log
3276 directory problem" with a 421 code for SMTP messages.
3278 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3279 0, 1, and 2 exist, because some OS complain for execs without them (see
3280 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3281 descriptors. However, the code omitted to check that the open succeeded,
3282 causing mysterious errors if for some reason the permissions on /dev/null
3283 got screwed. Now Exim writes a message to the main and panic logs, and
3284 bombs out if it can't open /dev/null.
3286 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3287 interact so that it is all more flexible. It is supposed to remain
3288 backwards compatible. Also added extra_local_interfaces.
3290 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3291 to bomb out with an assertion failure - to the client this appears as a
3292 connection drop. This problem occurs in the part of the code that was taken
3293 from the Samba project. Fortunately, the assertion is in a very simple
3294 function, so I have fixed this by reproducing the function inline in the
3295 one place where it is called, and arranging for authentication to fail
3296 instead of killing the process with assert().
3298 75. The SPA client code was not working when the server requested OEM rather
3299 than Unicode encoding.
3301 76. Added code to make require_files with a specific uid setting more usable in
3302 the case where statting the file as root fails - usually a non-root-mounted
3303 NFS file system. When this happens and the failure is EACCES, Exim now
3304 forks a subprocess and does the per-uid checking as the relevant uid.
3306 77. Added process_log_path.
3308 78. If log_file_path was not explicitly set, a setting of check_log_space or
3309 check_log_inodes was ignored.
3311 79. If a space check for the spool or log partitions fails, the incident is now
3312 logged. Of course, in the latter case the data may get lost...
3314 80. Added the %p formatting code to string_format() so that it can be used to
3315 print addresses in debug_print(). Adjusted all the address printing in the
3316 debugging in store.c to use %p rather than %d.
3318 81. There was a concern that a line of code in smtp_in.c could overflow a
3319 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3320 initially expressed, the concern was not well-founded, because trailing
3321 spaces are removed early. However, if the trailing spaces were followed by
3322 a NULL, they did not get removed, so the overflow was possible. Two fixes
3325 (a) I re-wrote the offending code in a cleaner fashion.
3326 (b) If an incoming SMTP command contains a NULL character, it is rejected
3329 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3330 runs initgroups(), so that if the Exim user is in any additional groups,
3331 they will be used during message reception.
3337 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob