1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.188 2005/08/01 15:01:12 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.
35 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
37 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
38 command itself ended in error, the underlying message about the transport
39 filter timeout was being overwritten with the pipe command error. Now the
40 underlying error message should be appended to the second error message.
42 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
44 PH/02 Reduce the amount of output that "make" produces by default. Full output
45 can still be requested.
47 PH/03 The warning log line about a condition test deferring for a "warn" verb
48 was being output only once per connection, rather than after each
49 occurrence (because it was using the same function as for successful
50 "warn" verbs). This seems wrong, so I have changed it.
52 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
53 it should not have, which might have caused a crash in the right
54 circumstances, but probably never did.
56 PH/04 Installed a modified version of Tony Finch's patch to make submission
57 mode fix the return path as well as the Sender: header line, and to
58 add a /name= option so that you can make the user's friendly name appear
61 TF/03 Added the control = fakedefer ACL modifier.
63 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
64 Mark Lowes for thorough testing.
66 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
68 TK/03 Merged latest SRS patch from Miles Wilton.
70 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
71 with the definition in sysexits.h (which is #included earlier).
72 Fortunately, Exim does not actually use EX_OK. The code used to try to
73 preserve the sysexits.h value, by assumimg that macro definitions were
74 scanned for macro replacements. I have been disabused of this notion,
75 so now the code just undefines EX_OK before #including unistd.h.
77 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
78 in the smtp transport. When a block could not be written in a single
79 write() function, the timeout was being re-applied to each part-write.
80 This seems wrong - if the receiver was accepting one byte at a time it
81 would take for ever. The timeout is now adjusted when this happens. It
82 doesn't have to be particularly precise.
84 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
85 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
87 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
88 without a domain if the check to <postmaster@domain> fails.
90 SC/01 Eximstats: added -xls and the ability to specify output files
91 (patch written by Frank Heydlauf).
93 SC/02 Eximstats: use FileHandles for outputing results.
95 SC/03 Eximstats: allow any combination of xls, txt, and html output.
97 SC/04 Eximstats: fixed display of large numbers with -nvr option
99 SC/05 Eximstats: fixed merging of reports with empty tables.
101 SC/06 Eximstats: added the -include_original_destination flag
103 SC/07 Eximstats: removed tabs and trailing whitespace.
105 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
107 TK/06 MBOX spool code: Add real "From " MBOX separator line
108 so the .eml file is really in mbox format (even though
109 most programs do not really care). Patch from Alex Miller.
111 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
112 The latter is generated from $received_to and is only set if the
113 message has one envelope recipient. SA can use these headers,
114 obviously out-of-the-box. Patch from Alex Miller.
116 PH/08 The ${def test on a variable was returning false if the variable's
117 value was "0", contrary to what the specification has always said!
118 The result should be true unless the variable is empty.
120 PH/09 The syntax error of a character other than { following "${if
121 def:variable_name" (after optional whitespace) was not being diagnosed.
122 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
123 accidental colon was present, for example, could give incorrect results.
125 PH/10 Tidied the code in a number of places where the st_size field of a stat()
126 result is used (not including appendfile, where other changes are about
129 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
130 This involved changing a lot of size variables from int to off_t. It
131 should work with maildirs and everything.
133 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
134 spamd dying while we are connected to it.
136 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
139 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
140 function for BDB 4.3.
142 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
144 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
145 the sources for PCRE have been re-arranged and more files are now
148 PH/15 The code I had for printing potentially long long variables in PH/11
149 above was not the best (it lost precision). The length of off_t variables
150 is now inspected at build time, and an appropriate printing format (%ld
151 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
152 to be "long long int" or "long int". This is needed for the internal
153 formatting function string_vformat().
155 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
156 the configuration file to be ":syslog", then the script "guesses" where
157 the logs files are, rather than using the compiled in default. In our
158 case the guess is not the same as the compiled default, so the script
159 suddenly stopped working when I started to use syslog. The patch checks
160 to see if log_file_path is "". If so, it attempts to read it from exim
161 with no configuration file to get the compiled in version, before it
162 falls back to the previous guessing code."
164 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
165 implementing BATV in an Exim configuration. See NewStuff for the gory
168 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
169 Makefile that are specific to HP-UX.
171 PH/18 If the "use_postmaster" option was set for a recipient callout together
172 with the "random" option, the postmaster address was used as the MAIL
173 FROM address for the random test, but not for the subsequent recipient
174 test. It is now used for both.
176 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
177 patch removes a few documentation additions to RFC 3028, because the
178 latest draft now contains them. It adds the new en;ascii-case comparator
179 and a new error check for 8bit text in MIME parts. Comparator and
180 require names are now matched exactly. I enabled the subaddress
181 extension, but it is not well tested yet (read: it works for me)."
183 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
184 rework some of the code of TK/09 above to avoid the hardwired use of
185 "%lld" and "long long". Replaced the call to snprintf() with a call to
188 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
189 records point to non-existent hosts", "retry timeout exceeded", and
190 "retry time not reached for any host after a long failure period".
192 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
193 experimental DomainKeys support:
195 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
196 (2) On an error such as an illegally used "control", the wrong name for
197 the control was given.
199 These problems did NOT occur unless DomainKeys support was compiled.
201 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
203 PH/24 Added ${if match_ip condition.
205 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
206 databases so that it will be absolutely obvious if a crash occurs in the
207 DB library. This is a regular occurrence (often caused by mis-matched
210 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
211 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
212 picked up on a user's system that detects such things. There doesn't seem
213 to be a gcc warning option for this - only an attribute that has to be
214 put on the function's prototype. It seems that in Fedora Core 4 they have
215 set this on a number of new functions. No doubt there will be more in due
218 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
219 specify a transport. However, if an address that was verified by such a
220 router was the subject of a callout, Exim crashed because it tried to
221 read the rcpt_include_affixes from the non-existent transport. Now it
222 just assumes that the setting of that option is false. This bug was
223 introduced by 4.51/PH/31.
225 PH/28 Changed -d+all to exclude +memory, because that information is very
226 rarely of interest, but it makes the output a lot bigger. People tend to
227 do -d+all out of habit.
229 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
230 code in os-type was giving problems when libc.so lives in lib64, like on
233 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
234 aren't the modern standard, and it seems that some systems' include files
235 don't always have them. Exim was already checking for some of the newer
236 ones like T_AAAA, and defining it itself. I've added checks for all the
237 record types that Exim uses.
239 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
240 not automatically generating a new one, as it is supposed to. This
241 prevented TLS from working. If the file did exist, but contained invalid
242 data, a new version was generated, as expected. It was only the case of a
243 non-existent file that was broken.
245 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
246 with a change in libdomainkeys > 0.64.
248 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
249 from DNS. If the selector record carries the flag, it now has
250 precedence over the domain-wide flag.
252 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
254 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
255 the use of a port name that isn't defined in /etc/services) occurred, the
256 message was deferred as in a normal delivery, and thus remained on the
257 spool, instead of being failed because of the mua_wrapper setting. This
258 is now fixed, and I tidied up some of the mua_wrapper messages at the
261 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
262 the messages in a hash of arrays rather than using individual hashes.
263 This is a bit cleaner and results in dramatic memory savings, albeit
264 at a slight CPU cost.
266 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
267 as requested by Marc Sherman.
269 SC/10 Eximstats: added histograms for user specified patterns as requested
272 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
274 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
275 fopen() in the content-scanning modules that did not already have it.
281 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
282 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
284 TK/02 Fix ACL "control" statement not being available in MIME ACL.
286 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
288 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
289 to test Sieve filters that use "vacation".
291 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
292 that changes the way the GnuTLS parameters are stored in the cache file.
293 The new format can be generated externally. For backward compatibility,
294 if the data in the cache doesn't make sense, Exim assumes it has read an
295 old-format file, and it generates new data and writes a new file. This
296 means that you can't go back to an older release without removing the
299 PH/03 A redirect router that has both "unseen" and "one_time" set does not
300 work if there are any delivery delays because "one_time" forces the
301 parent to be marked "delivered", so its unseen clone is never tried
302 again. For this reason, Exim now forbids the simultaneous setting of
305 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
306 redirected to themselves ("homonym" addresses). Read the long ChangeLog
307 entry if you want to know the details. The fix, however, neglected to
308 consider the case when local delivery batching is involved. The test for
309 "previously delivered" was not happening when checking to see if an
310 address could be batched with a previous (undelivered) one; under
311 certain circumstances this could lead to multiple deliveries to the same
314 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
315 in its include files, and this causes problems building Exim.
317 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
318 header_syntax) but Exim was just ignoring anything given after a slash.
319 In particular, this caused confusion with an attempt to use "verify =
320 reverse_host_lookup/defer_ok". An error is now given when options are
321 supplied for verify items that do not have them. (Maybe reverse_host_
322 lookup should have a defer_ok option, but that's a different point.)
324 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
325 defined by RFC 821) to 2048, because there were problems with some AUTH
326 commands, and RFC 1869 says the size should be increased for extended
327 SMTP commands that take arguments.
329 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
332 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
333 "unknown" error; now it says that the functionality isn't in the binary.
335 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
336 an address' error message when a string expansion fails (syntax or
337 whatever). Otherwise the password may appear in the log. Following change
338 PH/42 below, there is no longer a chance of it appearing in a bounce
341 PH/11 Installed exipick version 20050225.0 from John Jetmore.
343 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
344 of its addresses was ever tried. (Bugzilla bug #2.)
346 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
347 the result incorrectly in the debug output. (It correctly added a newline
348 to what was transported.)
350 TF/01 Added $received_time.
352 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
353 commented out examples of how to interface to a virus scanner and to
354 SpamAssassin. Also added commented examples of av_scanner and
355 spamd_address settings.
357 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
358 and controls are allowed in which ACLs. There were a couple of minor
359 errors. Some of the entries in the conditions table (which is a table of
360 where they are NOT allowed) were getting very unwieldy; rewrote them as a
361 negation of where the condition IS allowed.
363 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
365 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
366 header file does not have a version number, so I've had to invent a new
367 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
368 API. The code is untested by me (my Linux distribution still has 0.3.2 of
369 radiusclient), but it was contributed by a Radius user.
371 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
372 files or directories, for OpenSSL.
374 PH/19 When an Exim process that is running as root has to create an Exim log
375 file, it does so in a subprocess that runs as exim:exim so as to get the
376 ownership right at creation (otherwise, other Exim processes might see
377 the file with the wrong ownership). There was no test for failure of this
378 fork() call, which would lead to the process getting stuck as it waited
379 for a non-existent subprocess. Forks do occasionally fail when resources
380 run out. I reviewed all the other calls to fork(); they all seem to check
383 PH/20 When checking for unexpected SMTP input at connect time (before writing
384 the banner), Exim was not dealing correctly with a non-positive return
385 from the read() function. If the client had disconnected by this time,
386 the result was a log entry for a synchronization error with an empty
387 string after "input=" when read() returned zero. If read() returned -1
388 (an event I could not check), uninitialized data bytes were printed.
389 There were reports of junk text (parts of files, etc) appearing after
392 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
394 PH/22 Added support for macro redefinition, and (re)definition in between
395 driver and ACL definitions.
397 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
398 forgetting to use the resulting value; it was using the unexpanded value.
400 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
401 hadn't been configured. The fix is from Juergen Kreileder, who
402 understands it better than I do:
404 "Here's what I see happening with three configured cyrus_sasl
405 authenticators configured (plain, login, cram-md5):
407 On startup auth_cyrus_sasl_init() gets called for each of these.
408 This means three calls to sasl_listmech() without a specified mech_list.
409 => SASL tests which mechs of all available mechs actually work
410 => three warnings about OTP not working
411 => the returned list contains: plain, login, cram-md5, digest-md5, ...
413 With the patch, sasl_listmech() also gets called three times. But now
414 SASL's mech_list option is set to the server_mech specified in the the
415 authenticator. Or in other words, the answer from sasl_listmech()
416 gets limited to just the mech you're testing for (which is different
418 => the return list contains just 'plain' or 'login', 'cram-md5' or
419 nothing depending on the value of ob->server_mech.
421 I've just tested the patch: Authentication still works fine,
422 unavailable mechs specified in the exim configuration are still
423 caught, and the auth.log warnings about OTP are gone."
425 PH/25 When debugging is enabled, the contents of the command line are added
426 to the debugging output, even when log_selector=+arguments is not
429 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
430 answer is "GNU", and only if the return is "GNU/something" is the answer
433 PH/27 $acl_verify_message is now set immediately after the failure of a
434 verification in an ACL, and so is available in subsequent modifiers. In
435 particular, the message can be preserved by coding like this:
437 warn !verify = sender
438 set acl_m0 = $acl_verify_message
440 Previously, $acl_verify_message was set only while expanding "message"
441 and "log_message" when a very denied access.
443 PH/28 Modified OS/os.c-Linux with
445 -#ifndef OS_LOAD_AVERAGE
446 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
448 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
449 nomenclature these days.)
451 PH/29 Installed patch from the Sieve maintainer that adds the options
452 sieve_useraddress and sieve_subaddress to the redirect router.
454 PH/30 In these circumstances:
455 . Two addresses routed to the same list of hosts;
456 . First host does not offer TLS;
457 . First host accepts first address;
458 . First host gives temporary error to second address;
459 . Second host offers TLS and a TLS session is established;
460 . Second host accepts second address.
461 Exim incorrectly logged both deliveries with the TLS parameters (cipher
462 and peerdn, if requested) that were in fact used only for the second
465 PH/31 When doing a callout as part of verifying an address, Exim was not paying
466 attention to any local part prefix or suffix that was matched by the
467 router that accepted the address. It now behaves in the same way as it
468 does for delivery: the affixes are removed from the local part unless
469 rcpt_include_affixes is set on the transport.
471 PH/32 Add the sender address, as F=<...>, to the log line when logging a
472 timeout during the DATA phase of an incoming message.
474 PH/33 Sieve envelope tests were broken for match types other than :is. I have
475 applied a patch sanctioned by the Sieve maintainer.
477 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
478 the uid or gid is negative. A case of a negative gid caused this to be
479 noticed. The fix allows for either to be negative.
481 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
482 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
485 PH/36 The change PH/12 above was broken. Fixed it.
487 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
488 the grounds that routing the same address twice would always produce the
489 same answer. This might have been true once, but it is certainly no
490 longer true now. Routing a child address may depend on the previous
491 routing that produced that child. Some complicated redirection strategies
492 went wrong when messages had multiple recipients, and made Exim's
493 behaviour dependent on the order in which the addresses were given.
495 I have moved the duplicate checking until after the routing is complete.
496 Exim scans the addresses that are assigned to local and remote
497 transports, and removes any duplicates. This means that more work will be
498 done, as duplicates will always all be routed, but duplicates are
499 presumably rare, so I don't expect this is of any significance.
501 For deliveries to pipes, files, and autoreplies, the duplicate checking
502 still happens during the routing process, since they are not going to be
505 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
506 It corrects a timeout issue with spamd. This is Ian's comment: "The
507 background is that sometimes spamd either never reads data from a
508 connection it has accepted, or it never writes response data. The exiscan
509 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
510 blindly assumes that writes won't block so it may never time out."
512 PH/39 Allow G after quota size as well as K and M.
514 PH/40 The value set for $authenticated_id in an authenticator may not contain
515 binary zeroes or newlines because the value is written to log lines and
516 to spool files. There was no check on this. Now the value is run through
517 the string_printing() function so that such characters are converted to
518 printable escape sequences.
520 PH/41 $message_linecount is a new variable that contains the total number of
521 lines in the message. Compare $body_linecount, which is the count for the
524 PH/42 Exim no longer gives details of delivery errors for specific addresses in
525 bounce and delay warning messages, except in certain special cases, which
528 (a) An SMTP error message from a remote host;
529 (b) A message specified in a :fail: redirection;
530 (c) A message specified in a "fail" command in a system filter;
531 (d) A message specified in a FAIL return from the queryprogram router;
532 (e) A message specified by the cannot_route_message router option.
534 In these cases only, Exim does include the error details in bounce and
535 warning messages. There are also a few cases where bland messages such
536 as "unrouteable address" or "local delivery error" are given.
538 PH/43 $value is now also set for the "else" part of a ${run expansion.
540 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
541 being worked on, but at least Exim now implements the latest version to
544 PH/45 In a pipe transport, although a timeout while waiting for the pipe
545 process to complete was treated as a delivery failure, a timeout while
546 writing the message to the pipe was logged, but erroneously treated as a
547 successful delivery. Such timeouts include transport filter timeouts. For
548 consistency with the overall process timeout, these timeouts are now
549 treated as errors, giving rise to delivery failures by default. However,
550 there is now a new Boolean option for the pipe transport called
551 timeout_defer, which, if set TRUE, converts the failures into defers for
552 both kinds of timeout. A transport filter timeout is now identified in
555 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
556 systems where "make" and "gmake" are different, calling "gmake" at top
557 level broke things. I've arranged for the value of $(MAKE) to be passed
558 from the Makefile to this script so that it can call the same version of
562 A note about Exim versions 4.44 and 4.50
563 ----------------------------------------
565 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
566 changes of various kinds. As a consequence, a big documentation update was
567 needed. This delayed the release for rather longer than seemed good, especially
568 in the light of a couple of (minor) security issues. Therefore, the changes
569 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
570 release. So 4.44 and 4.50 are in effect two different branches that both start
573 I have left the 4.50 change log unchanged; it contains all the changes since
574 4.43. The change log for 4.44 is below; many of its items are identical to
575 those for 4.50. This seems to be the most sensible way to preserve the
576 historical information.
582 1. Minor wording change to the doc/README.SIEVE file.
584 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
585 computation of the current number of files was incorrect.
587 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
588 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
591 4. Give more explanation in the error message when the command for a transport
592 filter fails to execute.
594 5. There are several places where Exim runs a non-Exim command in a
595 subprocess. The SIGUSR1 signal should be disabled for these processes. This
596 was being done only for the command run by the queryprogram router. It is
597 now done for all such subprocesses. The other cases are: ${run, transport
598 filters, and the commands run by the lmtp and pipe transports.
600 6. Added CONFIGURE_GROUP build-time option.
602 7. Some older OS have a limit of 256 on the maximum number of file
603 descriptors. Exim was using setrlimit() to set 1000 as a large value
604 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
605 systems. I've change it so that if it can't get 1000, it tries for 256.
607 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
608 was an oversight, and furthermore, ever since the addition of extra
609 controls (e.g. 4.43/32), the checks on when to allow different forms of
610 "control" were broken. There should now be diagnostics for all cases when a
611 control that does not make sense is encountered.
613 9. Added the /retain_sender option to "control=submission".
615 10. $recipients is now available in the predata ACL (oversight).
617 11. Tidy the search cache before the fork to do a delivery from a message
618 received from the command line. Otherwise the child will trigger a lookup
619 failure and thereby defer the delivery if it tries to use (for example) a
620 cached ldap connection that the parent has called unbind on.
622 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
623 of $address_data from the recipient verification was clobbered by the
626 13. The value of address_data from a sender verification is now available in
627 $sender_address_data in subsequent conditions in the ACL statement.
629 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
631 15. Added a new option "connect=<time>" to callout options, to set a different
634 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
635 was its contents. (It was OK if the option was not defined at all.)
637 17. A "Completed" log line is now written for messages that are removed from
638 the spool by the -Mrm option.
640 18. New variables $sender_verify_failure and $recipient_verify_failure contain
641 information about exactly what failed.
643 19. Added -dd to debug only the daemon process.
645 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
646 handles timeouts, both on the server side and network timeouts. Renamed the
647 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
649 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
650 It is now set to "smtps".
652 22. $host_address is now set to the target address during the checking of
655 23. When checking ignore_target_hosts for an ipliteral router, no host name was
656 being passed; this would have caused $sender_host_name to have been used if
657 matching the list had actually called for a host name (not very likely,
658 since this list is usually IP addresses). A host name is now passed as
661 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
662 code that specifies a non-restarting handler (typically sigaction() in
663 modern systems) in an attempt to fix a rare and obscure crash bug.
665 25. Narrowed the window for a race in the daemon that could cause it to ignore
666 SIGCHLD signals. This is not a major problem, because they are used only to
667 wake it up if nothing else does.
669 26. A malformed maildirsize file could cause Exim to calculate negative values
670 for the mailbox size or file count. Odd effects could occur as a result.
671 The maildirsize information is now recalculated if the size or filecount
674 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
675 support for a long time. Removed HAVE_SYS_VFS_H.
677 28. Installed the latest version of exipick from John Jetmore.
679 29. In an address list, if the pattern was not a regular expression, an empty
680 subject address (from a bounce message) matched only if the pattern was an
681 empty string. Non-empty patterns were not even tested. This was the wrong
682 because it is perfectly reasonable to use an empty address as part of a
683 database query. An empty address is now tested by patterns that are
684 lookups. However, all the other forms of pattern expect the subject to
685 contain a local part and a domain, and therefore, for them, an empty
686 address still always fails if the pattern is not itself empty.
688 30. Exim went into a mad DNS loop when attempting to do a callout where the
689 host was specified on an smtp transport, and looking it up yielded more
692 31. Re-factored the code for checking spool and log partition space into a
693 function that finds that data and another that does the check. The former
694 is then used to implement four new variables: $spool_space, $log_space,
695 $spool_inodes, and $log_inodes.
697 32. The RFC2047 encoding function was originally intended for short strings
698 such as real names; it was not keeping to the 75-character limit for
699 encoded words that the RFC imposes. It now respects the limit, and
700 generates multiple encoded words if necessary. To be on the safe side, I
701 have increased the buffer size for the ${rfc2047: expansion operator from
704 33. It is now permitted to omit both strings after an "if" condition; if the
705 condition is true, the result is "true". As before, when the second string
706 is omitted, a false condition yields an empty string. This makes it less
707 cumbersome to write custom ACL and router conditions.
709 34. Failure to deliver a bounce message always caused it to be frozen, even if
710 there was an errors_to setting on the router. The errors_to setting is now
713 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
714 canonical form (fully expanded) before being placed in
715 $sender_host_address.
717 36. The table in the code that translates DNS record types into text (T_A to
718 "A" for instance) was missing entries for NS and CNAME. It is just possible
719 that this could have caused confusion if both these types were looked up
720 for the same domain, because the text type is used as part of Exim's
721 per-process caching. But the chance of anyone hitting this buglet seems
724 37. The dnsdb lookup has been extended in a number of ways.
726 (1) There is a new type, "zns", which walks up the domain tree until it
727 finds some nameserver records. It should be used with care.
729 (2) There is a new type, "mxh", which is like "mx" except that it returns
730 just the host names, not the priorities.
732 (3) It is now possible to give a list of domains (or IP addresses) to be
733 looked up. The behaviour when one of the lookups defers can be
734 controlled by a keyword.
736 (4) It is now possible to specify the separator character for use when
737 multiple records are returned.
739 38. The dnslists ACL condition has been extended: it is now possible to supply
740 a list of IP addresses and/or domains to be looked up in a particular DNS
743 39. Added log_selector=+queue_time_overall.
745 40. When running the queue in the test harness, wait just a tad after forking a
746 delivery process, to get repeatability of debugging output.
748 41. Include certificate and key file names in error message when GnuTLS fails
749 to set them up, because the GnuTLS error message doesn't include the name
750 of the failing file when there is a problem reading it.
752 42. Allow both -bf and -bF in the same test run.
754 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
756 44. The "Exiscan patch" is now merged into the mainline Exim source.
758 45. Sometimes the final signoff response after QUIT could fail to get
759 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
760 < 0 before doing a fflush(). This bug looks as though it goes back to the
761 introduction of TLS in release 3.20, but "sometimes" must have been rare
762 because the tests only now provoked it.
764 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
765 (this can affect the format of dates).
767 47. exim_tidydb, when checking for the continued existence of a message for
768 which it has found a message-specific retry record, was not finding
769 messages that were in split spool directories. Consequently, it was
770 deleting retry records that should have stayed in existence.
772 48. Steve fixed some bugs in eximstats.
774 49. The SPA authentication driver was not abandoning authentication and moving
775 on to the next authenticator when an expansion was forced to fail,
776 contradicting the general specification for all authenticators. Instead it
777 was generating a temporary error. It now behaves as specified.
779 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
780 (the order specifies the preference for clients). The order is now AES256,
781 AES128, 3DES, ARCFOUR128.
783 51. Small patch to Sieve code - explicitly set From: when generating an
786 52. Exim crashed if a remote delivery caused a very long error message to be
787 recorded - for instance if somebody sent an entire SpamAssassin report back
788 as a large number of 550 error lines. This bug was coincidentally fixed by
789 increasing the size of one of Exim's internal buffers (big_buffer) that
790 happened as part of the Exiscan merge. However, to be on the safe side, I
791 have made the code more robust (and fixed the comments that describe what
794 53. Now that there can be additional text after "Completed" in log lines (if
795 the queue_time_overall log selector is set), a one-byte patch to exigrep
796 was needed to allow it to recognize "Completed" as not the last thing in
799 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
800 patch that reportedly fixes this has been added. I am not expert enough to
801 create a test for it. This is what the patch creator wrote:
803 "I found a little strange behaviour of ldap code when working with
804 Windows 2003 AD Domain, where users was placed in more than one
805 Organization Units. When I tried to give exim partial DN, the exit code
806 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
807 But simultaneously result of request was absolutely normal ldap result,
808 so I produce this patch..."
810 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
811 REFERENCE, so I have modified the code to exclude the patch when that macro
814 55. Some experimental protocols are using DNS PTR records for new purposes. The
815 keys for these records are domain names, not reversed IP addresses. The
816 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
817 leaves it alone. Component reversal etc. now happens only for IP addresses.
820 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
822 57. Double the size of the debug message buffer (to 2048) so that more of very
823 long debug lines gets shown.
825 58. The exicyclog utility now does better if the number of log files to keep
826 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
828 59. Two changes related to the smtp_active_hostname option:
830 (1) $smtp_active_hostname is now available as a variable.
831 (2) The default for smtp_banner uses $smtp_active_hostname instead
832 of $primary_hostname.
834 60. The host_aton() function is supposed to be passed a string that is known
835 to be a valid IP address. However, in the case of IPv6 addresses, it was
836 not checking this. This is a hostage to fortune. Exim now panics and dies
837 if the condition is not met. A case was found where this could be provoked
838 from a dnsdb PTR lookup with an IPv6 address that had more than 8
839 components; fortuitously, this particular loophole had already been fixed
840 by change 4.50/55 above.
842 If there are any other similar loopholes, the new check in host_aton()
843 itself should stop them being exploited. The report I received stated that
844 data on the command line could provoke the exploit when Exim was running as
845 exim, but did not say which command line option was involved. All I could
846 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
850 61. There was a buffer overflow vulnerability in the SPA authentication code
851 (which came originally from the Samba project). I have added a test to the
852 spa_base64_to_bits() function which I hope fixes it.
855 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
856 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
858 63. The daemon start-up calls getloadavg() while still root for those OS that
859 need the first call to be done as root, but it missed one case: when
860 deliver_queue_load_max is set with deliver_drop_privilege. This is
861 necessary for the benefit of the queue runner, because there is no re-exec
862 when deliver_drop_privilege is set.
864 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
867 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
868 was in use, was not putting the data itself into the right store pool;
869 consequently, it could be overwritten for a subsequent message in the same
870 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
873 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
875 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
876 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
877 the function were treating the return as a boolean value, which happened to
878 work because 0=false and not-0=true, but is not correct code.
880 68. The host_aton() function was not handling scoped IPv6 addresses (those
881 with, for example, "%eth0" on the end) correctly.
883 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
884 negated items (that is, ~something) in unsigned ints. Some compilers
885 apparently mutter when there is no cast.
887 70. If an address verification called from an ACL failed, and did not produce a
888 user-specific message (i.e. there was only a "system" message), nothing was
889 put in $acl_verify_message. In this situation, it now puts the system
892 71. Change 4.23/11 added synchronization checking at the start of an SMTP
893 session; change 4.31/43 added the unwanted input to the log line - except
894 that it did not do this in the start of session case. It now does.
896 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
897 This is wrong and can cause the other end to generate a synchronization
898 error if it is another Exim or anything else that does the synchronization
899 check. A QUIT command is no longer sent after a timeout.
901 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
904 74. The defer_ok option of callout verification was not working if it was used
905 when verifying addresses in header lines, that is, for this case:
907 verify = header_sender/callout=defer_ok
909 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
910 those file descriptors could be used for SMTP connections. If anything
911 wrote to stderr (the example that came up was "warn" in embedded Perl), it
912 could be sent to the SMTP client, causing chaos. The daemon now opens
913 stdin, stdout, and stderr to /dev/null when it puts itself into the
916 76. Arrange for output from Perl's "warn" command to be written to Exim's main
917 log by default. The user can override this with suitable Perl magic.
919 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
920 the log message when discard triggers, was not working for the DATA ACL or
921 for the non-SMTP ACL.
923 78. Error message wording change in sieve.c.
925 79. If smtp_accept_max_per_host was set, the number of connections could be
926 restricted to fewer than expected, because the daemon was trying to set up
927 a new connection before checking whether the processes handling previous
928 connections had finished. The check for completed processes is now done
929 earlier. On busy systems, this bug wouldn't be noticed because something
930 else would have woken the daemon, and it would have reaped the completed
933 80. If a message was submitted locally by a user whose login name contained one
934 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
935 It caused a spool format error. I have fixed the spool reading code. A
936 related problem was that the "from" clause in the Received: line became
937 illegal because of the space(s). It is now covered by ${quote_local_part.
939 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
942 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
944 83. Patch from Sieve maintainer to fix unterminated string problem in
947 84. Some minor changes to the Linux configuration files to help with other
948 OS variants using glibc.
950 85. One more patch for Sieve to update vacation handling to latest spec.
953 ----------------------------------------------------
954 See the note above about the 4.44 and 4.50 releases.
955 ----------------------------------------------------
961 1. Change 4.43/35 introduced a bug that caused file counts to be
962 incorrectly computed when quota_filecount was set in an appendfile
965 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
966 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
969 3. Give more explanation in the error message when the command for a transport
970 filter fails to execute.
972 4. There are several places where Exim runs a non-Exim command in a
973 subprocess. The SIGUSR1 signal should be disabled for these processes. This
974 was being done only for the command run by the queryprogram router. It is
975 now done for all such subprocesses. The other cases are: ${run, transport
976 filters, and the commands run by the lmtp and pipe transports.
978 5. Some older OS have a limit of 256 on the maximum number of file
979 descriptors. Exim was using setrlimit() to set 1000 as a large value
980 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
981 systems. I've change it so that if it can't get 1000, it tries for 256.
983 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
984 was an oversight, and furthermore, ever since the addition of extra
985 controls (e.g. 4.43/32), the checks on when to allow different forms of
986 "control" were broken. There should now be diagnostics for all cases when a
987 control that does not make sense is encountered.
989 7. $recipients is now available in the predata ACL (oversight).
991 8. Tidy the search cache before the fork to do a delivery from a message
992 received from the command line. Otherwise the child will trigger a lookup
993 failure and thereby defer the delivery if it tries to use (for example) a
994 cached ldap connection that the parent has called unbind on.
996 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
997 of $address_data from the recipient verification was clobbered by the
1000 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1001 was its contents. (It was OK if the option was not defined at all.)
1003 11. A "Completed" log line is now written for messages that are removed from
1004 the spool by the -Mrm option.
1006 12. $host_address is now set to the target address during the checking of
1007 ignore_target_hosts.
1009 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1010 being passed; this would have caused $sender_host_name to have been used if
1011 matching the list had actually called for a host name (not very likely,
1012 since this list is usually IP addresses). A host name is now passed as
1015 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1016 code that specifies a non-restarting handler (typically sigaction() in
1017 modern systems) in an attempt to fix a rare and obscure crash bug.
1019 15. Narrowed the window for a race in the daemon that could cause it to ignore
1020 SIGCHLD signals. This is not a major problem, because they are used only to
1021 wake it up if nothing else does.
1023 16. A malformed maildirsize file could cause Exim to calculate negative values
1024 for the mailbox size or file count. Odd effects could occur as a result.
1025 The maildirsize information is now recalculated if the size or filecount
1028 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1029 support for a long time. Removed HAVE_SYS_VFS_H.
1031 18. Updated exipick to current release from John Jetmore.
1033 19. Allow an empty sender to be matched against a lookup in an address list.
1034 Previously the only cases considered were a regular expression, or an
1037 20. Exim went into a mad DNS lookup loop when doing a callout where the
1038 host was specified on the transport, if the DNS lookup yielded more than
1041 21. The RFC2047 encoding function was originally intended for short strings
1042 such as real names; it was not keeping to the 75-character limit for
1043 encoded words that the RFC imposes. It now respects the limit, and
1044 generates multiple encoded words if necessary. To be on the safe side, I
1045 have increased the buffer size for the ${rfc2047: expansion operator from
1048 22. Failure to deliver a bounce message always caused it to be frozen, even if
1049 there was an errors_to setting on the router. The errors_to setting is now
1052 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1053 canonical form (fully expanded) before being placed in
1054 $sender_host_address.
1056 24. Updated eximstats to version 1.33
1058 25. Include certificate and key file names in error message when GnuTLS fails
1059 to set them up, because the GnuTLS error message doesn't include the name
1060 of the failing file when there is a problem reading it.
1062 26. Expand error message when OpenSSL has problems setting up cert/key files.
1065 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1066 (this can affect the format of dates).
1068 28. exim_tidydb, when checking for the continued existence of a message for
1069 which it has found a message-specific retry record, was not finding
1070 messages that were in split spool directories. Consequently, it was
1071 deleting retry records that should have stayed in existence.
1073 29. eximstats updated to version 1.35
1074 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1075 1.35 - bugfix such that pie charts by volume are generated correctly
1077 30. The SPA authentication driver was not abandoning authentication and moving
1078 on to the next authenticator when an expansion was forced to fail,
1079 contradicting the general specification for all authenticators. Instead it
1080 was generating a temporary error. It now behaves as specified.
1082 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1083 (the order specifies the preference for clients). The order is now AES256,
1084 AES128, 3DES, ARCFOUR128.
1086 31. Small patch to Sieve code - explicitly set From: when generating an
1089 32. Exim crashed if a remote delivery caused a very long error message to be
1090 recorded - for instance if somebody sent an entire SpamAssassin report back
1091 as a large number of 550 error lines. This bug was coincidentally fixed by
1092 increasing the size of one of Exim's internal buffers (big_buffer) that
1093 happened as part of the Exiscan merge. However, to be on the safe side, I
1094 have made the code more robust (and fixed the comments that describe what
1097 33. Some experimental protocols are using DNS PTR records for new purposes. The
1098 keys for these records are domain names, not reversed IP addresses. The
1099 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1100 leaves it alone. Component reversal etc. now happens only for IP addresses.
1103 34. The host_aton() function is supposed to be passed a string that is known
1104 to be a valid IP address. However, in the case of IPv6 addresses, it was
1105 not checking this. This is a hostage to fortune. Exim now panics and dies
1106 if the condition is not met. A case was found where this could be provoked
1107 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1108 components; fortuitously, this particular loophole had already been fixed
1109 by change 4.50/55 or 4.44/33 above.
1111 If there are any other similar loopholes, the new check in host_aton()
1112 itself should stop them being exploited. The report I received stated that
1113 data on the command line could provoke the exploit when Exim was running as
1114 exim, but did not say which command line option was involved. All I could
1115 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1116 running as the user.
1119 35. There was a buffer overflow vulnerability in the SPA authentication code
1120 (which came originally from the Samba project). I have added a test to the
1121 spa_base64_to_bits() function which I hope fixes it.
1124 36. The daemon start-up calls getloadavg() while still root for those OS that
1125 need the first call to be done as root, but it missed one case: when
1126 deliver_queue_load_max is set with deliver_drop_privilege. This is
1127 necessary for the benefit of the queue runner, because there is no re-exec
1128 when deliver_drop_privilege is set.
1130 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1131 was in use, was not putting the data itself into the right store pool;
1132 consequently, it could be overwritten for a subsequent message in the same
1133 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1136 38. Sometimes the final signoff response after QUIT could fail to get
1137 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1138 < 0 before doing a fflush(). This bug looks as though it goes back to the
1139 introduction of TLS in release 3.20, but "sometimes" must have been rare
1140 because the tests only now provoked it.
1146 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1147 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1148 Now it returns an integer. A number of calls were still expecting a T/F
1149 return. Fortuitously, in all cases, the tests worked in OK situations,
1150 which is the norm. However, things would have gone wrong on any write
1151 failures on the smtp file descriptor. This function is used when sending
1152 messages over SMTP and also when doing verify callouts.
1154 2. When Exim is called to do synchronous delivery of a locally submitted
1155 message (the -odf or -odi options), it no longer closes stderr before doing
1158 3. Implemented the mua_wrapper option.
1160 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1162 5. Implemented the functions header_remove(), header_testname(),
1163 header_add_at_position(), and receive_remove_recipient(), and exported them
1166 6. If an ACL "warn" statement specified the addition of headers, Exim already
1167 inserted X-ACL-Warn: at the start if there was no header name. However, it
1168 was not making this test for the second and subsequent header lines if
1169 there were newlines in the string. This meant that an invalid header could
1170 be inserted if Exim was badly configured.
1172 7. Allow an ACL "warn" statement to add header lines at the start or after all
1173 the Received: headers, as well as at the end.
1175 8. Added the rcpt_4xx retry error code.
1177 9. Added postmaster_mailfrom=xxx to callout verification option.
1179 10. Added mailfrom=xxxx to the callout verification option, for verify=
1182 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1183 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1185 12. Inserted some casts to stop certain compilers warning when using pointer
1186 differences as field lengths or precisions in printf-type calls (mostly
1187 affecting debugging statements).
1189 13. Added optional readline() support for -be (dynamically loaded).
1191 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1192 same clock tick as a message's arrival, so that its received time was the
1193 same as the "first fail" time on the retry record, and that message
1194 remained on the queue past the ultimate address timeout, every queue runner
1195 would try a delivery (because it was past the ultimate address timeout) but
1196 after another failure, the ultimate address timeout, which should have then
1197 bounced the address, did not kick in. This was a "< instead of <=" error;
1198 in most cases the first failure would have been in the next clock tick
1199 after the received time, and all would be well.
1201 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1202 being recognized when the domain list was tested by the match_domain
1203 condition in an expansion string.
1205 16. Added the ${str2b64: operator.
1207 17. Exim was always calling setrlimit() to set a large limit for the number of
1208 processes, without checking whether the existing limit was already
1209 adequate. (It did check for the limit on file descriptors.) Furthermore,
1210 errors from getrlimit() and setrlimit() were being ignored. Now they are
1211 logged to the main and panic logs, but Exim does carry on, to try to do its
1212 job under whatever limits there are.
1214 18. Imported PCRE 5.0.
1216 19. Trivial typo in log message " temporarily refused connection" (the leading
1219 20. If the log selector return_path_on_delivery was set and an address was
1220 redirected to /dev/null, the delivery process crashed because it assumed
1221 that a return path would always be set for a "successful" delivery. In this
1222 case, the whole delivery is bypassed as an optimization, and therefore no
1225 21. Internal re-arrangement: the function for sending a challenge and reading
1226 a response while authentication was assuming a zero-terminated challenge
1227 string. It's now changed to take a pointer and a length, to allow for
1228 binary data in such strings.
1230 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1232 23. Exim was not respecting finduser_retries when seeking the login of the
1233 uid under which it was called; it was always trying 10 times. (The default
1234 setting of finduser_retries is zero.) Also, it was sleeping after the final
1235 failure, which is pointless.
1237 24. Implemented tls_on_connect_ports.
1239 25. Implemented acl_smtp_predata.
1241 26. If the domain in control=submission is set empty, Exim assumes that the
1242 authenticated id is a complete email address when it generates From: or
1243 Sender: header lines.
1245 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1246 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1247 chown and chgrp in /bin and hostname in /usr/bin.
1249 28. Exim was keeping the "process log" file open after each use, just as it
1250 does for the main log. This opens the possibility of it remaining open for
1251 long periods when the USR1 signal hits a daemon. Occasional processlog
1252 errors were reported, that could have been caused by this. Anyway, it seems
1253 much more sensible not to leave this file open at all, so that is what now
1256 29. The long-running daemon process does not normally write to the log once it
1257 has entered its main loop, and it closes the log before doing so. This is
1258 so that log files can straightforwardly be renamed and moved. However,
1259 there are a couple of unusual error situations where the daemon does write
1260 log entries, and I had neglected to close the log afterwards.
1262 30. The text of an SMTP error response that was received during a remote
1263 delivery was being truncated at 512 bytes. This is too short for some of
1264 the long messages that one sometimes sees. I've increased the limit to
1267 31. It is now possible to make retry rules that apply only when a message has a
1268 specific sender, in particular, an empty sender.
1270 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1271 it possible to be selective about when SMTP synchronization is enforced.
1273 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1275 32. Implemented hosts_connection_nolog.
1277 33. Added an ACL for QUIT.
1279 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1282 35. Added mailbox_size and mailbox_filecount to appendfile.
1284 36. Added control = no_multiline_responses to ACLs.
1286 37. There was a bug in the logic of the code that waits for the clock to tick
1287 in the case where the clock went backwards by a substantial amount such
1288 that the microsecond fraction of "now" was more than the microsecond
1289 fraction of "then" (but the whole seconds number was less).
1291 38. Added support for the libradius Radius client library this is found on
1292 FreeBSD (previously only the radiusclient library was supported).
1298 1. When certain lookups returned multiple values in the form name=value, the
1299 quoting of the values was not always being done properly. Specifically:
1300 (a) If the value started with a double quote, but contained no whitespace,
1302 (b) If the value contained whitespace other than a space character (i.e.
1303 tabs or newlines or carriage returns) it was not quoted.
1304 This fix has been applied to the mysql and pgsql lookups by writing a
1305 separate quoting function and calling it from the lookup code. The fix
1306 should probably also be applied to nisplus, ibase and oracle lookups, but
1307 since I cannot test any of those, I have not disturbed their existing code.
1309 2. A hit in the callout cache for a specific address caused a log line with no
1310 reason for rejecting RCPT. Now it says "Previous (cached) callout
1311 verification failure".
1313 3. There was an off-by-one bug in the queryprogram router. An over-long
1314 return line was truncated at 256 instead of 255 characters, thereby
1315 overflowing its buffer with the terminating zero. As well as fixing this, I
1316 have increased the buffer size to 1024 (and made a note to document this).
1318 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1319 when Exim is waiting for an SMTP response from a remote server, Exim
1320 restarts its select() call on the socket, thereby resetting its timeout.
1321 This is not a problem when such interrupts are rare. Somebody set up a cron
1322 job to run exiwhat every 2 minutes, which is less than the normal select()
1323 timeout (5 or 10 minutes). This meant that the select() timeout never
1324 kicked in because it was always reset. I have fixed this by comparing the
1325 time when an interrupt arrives with the time at the start of the first call
1326 to select(). If more time than the timeout has elapsed, the interrupt is
1327 treated as a timeout.
1329 5. Some internal re-factoring in preparation for the addition of Sieve
1330 extensions (by MH). In particular, the "personal" test is moved to a
1331 separate function, and given an option for scanning Cc: and Bcc: (which is
1332 not set for Exim filters).
1334 6. When Exim created an email address using the login of the caller as the
1335 local part (e.g. when creating a From: or Sender: header line), it was not
1336 quoting the local part when it contained special characters such as @.
1338 7. Installed new OpenBSD configuration files.
1340 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1341 try to make them clearer.
1343 9. Callout options, other than the timeout value, were being ignored when
1344 verifying sender addresses in header lines. For example, when using
1346 verify = header_sender/callout=no_cache
1348 the cache was (incorrectly) being used.
1350 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1351 only the Cygwin environment.
1353 11. When return_path_on_delivery was set as a log selector, if different remote
1354 addresses in the same message used different return paths and parallel
1355 remote delivery occurred, the wrong values would sometimes be logged.
1356 (Whenever a remote delivery process finished, the return path value from
1357 the most recently started remote delivery process was logged.)
1359 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1360 lines when AUTH and/or TLS are in use. This is the "received protocol"
1361 field. Exim used to use "asmtp" for authenticated SMTP, without any
1362 indication (in the protocol name) for TLS use. Now it follows the RFC and
1363 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1364 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1365 names appear in log lines as well as in Received: header lines.
1367 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1368 extensions, and comparison tests, and to fix some bugs.
1370 14. Changes to the "personal" filter test:
1372 (1) The test was buggy in that it was just doing the equivalent of
1373 "contains" tests on header lines. For example, if a user's address was
1374 anne@some.where, the "personal" test would incorrectly be true for
1376 To: susanne@some.where
1378 This test is now done by extracting each address from the header in turn,
1379 and checking the entire address. Other tests that are part of "personal"
1380 are now done using regular expressions (for example, to check local parts
1381 of addresses in From: header lines).
1383 (2) The list of non-personal local parts in From: addresses has been
1384 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1385 taken from the Sieve specification recommendations.
1387 (3) If the message contains any header line starting with "List-" it is
1388 treated as non-personal.
1390 (4) The test for "circular" in the Subject: header line has been removed
1391 because it now seems ill-conceived.
1393 15. Minor typos in src/EDITME comments corrected.
1395 16. Installed latest exipick from John Jetmore.
1397 17. If headers_add on a router specified a text string that was too long for
1398 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1399 of string_sprintf() is now avoided.
1401 18. $message_body_size was not set (it was always zero) when running the DATA
1402 ACL and the local_scan() function.
1404 19. For the "mail" command in an Exim filter, no default was being set for
1405 the once_repeat time, causing a random time value to be used if "once" was
1406 specified. (If the value happened to be <= 0, no repeat happened.) The
1407 default is now 0s, meaning "never repeat". The "vacation" command was OK
1408 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1409 (I found it when inspecting the code).
1411 20. There is now an overall timeout for performing a callout verification. It
1412 defaults to 4 times the callout timeout, which applies to individual SMTP
1413 commands during the callout. The overall timeout applies when there is more
1414 than one host that can be tried. The timeout is checked before trying the
1415 next host. This prevents very long delays if there are a large number of
1416 hosts and all are timing out (e.g. when the network connections are timing
1417 out). The value of the overall timeout can be changed by specifying an
1418 additional sub-option for "callout", called "maxwait". For example:
1420 verify = sender/callout=5s,maxwait=20s
1422 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1423 to the end before writing, but this should make it even safer).
1425 22. Exim was forgetting that it had advertised PIPELINING for the second and
1426 subsequent messages on an SMTP connection. It was also not resetting its
1427 memory on STARTTLS and an internal HELO.
1429 23. When Exim logs an SMTP synchronization error within a session, it now
1430 records whether PIPELINING has been advertised or not.
1432 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1433 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1434 rather than long int.
1436 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1438 26. Added the never_mail option to autoreply.
1444 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1445 crash if the getsockname() call failed; this can happen if a connection is
1446 closed very soon after it is established. The problem was simply in the
1447 order in which certain operations were done, causing Exim to try to write
1448 to the SMTP stream before it had set up the file descriptor. The bug has
1449 been fixed by making things happen in the correct order.
1455 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1456 before the connection was closed, thus losing the rejection response.
1458 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1459 some early Solaris releases, but causes trouble in current releases where
1460 socklen_t is defined.
1462 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1465 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1468 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1469 that releasing the top of it at the end releases what was used for sub-
1470 expansions (unless the block got too big). However, discard this block if
1471 the first thing is a variable or header, so that we can use its block when
1472 it is dynamic (useful for very large $message_headers, for example).
1474 6. Lookups now cache *every* query, not just the most recent. A new, separate
1475 store pool is used for this. It can be recovered when all lookup caches are
1476 flushed. Lookups now release memory at the end of their result strings.
1477 This has involved some general refactoring of the lookup sources.
1479 7. Some code has been added to the store_xxx() functions to reduce the amount
1480 of flapping under certain conditions.
1482 8. log_incoming_interface used to affect only the <= reception log lines. Now
1483 it causes the local interface and port to be added to several more SMTP log
1484 lines, for example "SMTP connection from", and rejection lines.
1486 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1488 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1490 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1491 could be overwritten at the end of the current message (or the start of a
1492 new message if it was set in a HELO ACL). The value is now preserved for
1493 the duration of the SMTP connection.
1495 12. If a transport had a headers_rewrite setting, and a matching header line
1496 contained an unqualified address, that address was qualified, even if it
1497 did not match any rewriting rules. The underlying bug was that the values
1498 of the flags that permit the existence of unqualified sender and recipient
1499 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1500 non-local messages, and by -bnq for local messages) were not being
1501 preserved with the message after it was received.
1503 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1504 "next input=" as part of the text comprising the host identity instead of
1505 the correct text. The code was using the same buffer for two different
1506 strings. However, depending on which order the printing function evaluated
1507 its arguments, the bug did not always show up. Under Linux, for example, my
1508 test suite worked just fine.
1510 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1511 doesn't work with some older versions of Perl. It has been changed to "my",
1512 which in any case is probably the better facility to use.
1514 15. A really picky compiler found some instances of statements for creating
1515 error messages that either had too many or two few arguments for the format
1518 16. The size of the buffer for calls to the DNS resolver has been increased
1519 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1520 for addresses that have a lot of PTR records. This alleviates a problem; it
1521 does not fully solve it.
1523 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1524 buffer now truncates the list and logs the incident, which is the same
1525 action as happens when Exim is looking up a host name and its aliases.
1526 Previously in this situation something unpredictable would happen;
1527 sometimes it was "internal error: store_reset failed".
1529 18. If a server dropped the connection unexpectedly when an Exim client was
1530 using GnuTLS and trying to read a response, the client delivery process
1531 crashed while trying to generate an error log message.
1533 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1534 string, for example:
1536 warn message = H1: something\nH2: something
1538 the text was added as a single header line from Exim's point of view
1539 though it ended up OK in the delivered message. However, searching for the
1540 second and subsequent header lines using $h_h2: did not work. This has been
1541 fixed. Similarly, if a system filter added multiple headers in this way,
1542 the routers could not see them.
1544 20. Expanded the error message when iplsearch is called with an invalid key to
1545 suggest using net-iplsearch in a host list.
1547 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1548 ACLs are no longer actually imposed (and a message to that effect is
1551 22. If a "gecos" field in a passwd entry contained escaped characters, in
1552 particular, if it contained a \" sequence, Exim got it wrong when building
1553 a From: or a Sender: header from that name. A second bug also caused
1554 incorrect handling when an unquoted " was present following a character
1555 that needed quoting.
1557 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1558 was not being matched caselessly.
1560 24. Arranged for all hyphens in the exim.8 source to be escaped with
1563 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1564 properly. Recipient callout cache records were still being keyed to include
1565 the sender, even when use_sender was set false. This led to far more
1566 callouts that were necessary. The sender is no longer included in the key
1567 when use_sender is false.
1569 26. Added "control = submission" modifier to ACLs.
1571 27. Added the ${base62d: operator to decode base 62 numbers.
1573 28. dnsdb lookups can now access SRV records.
1575 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1576 the configuration file.
1578 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1579 (-v) mode. This makes the output for a verbose queue run more intelligible.
1581 31. Added a use_postmaster feature to recipient callouts.
1583 32. Added the $body_zerocount variable, containing the number of binary zero
1584 bytes in the message body.
1586 33. The time of last modification of the "new" subdirectory is now used as the
1587 "mailbox time last read" when there is a quota error for a maildir
1590 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1592 35. Added +ignore_unknown as a special item in host lists.
1594 36. Code for decoding IPv6 addresses in host lists is now included, even if
1595 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1596 address was recognized as an IP address, but was then not correctly decoded
1597 into binary, causing unexpected and incorrect effects when compared with
1604 1. Very minor rewording of debugging text in manualroute to say "list of
1605 hosts" instead of "hostlist".
1607 2. If verify=header_syntax was set, and a header line with an unqualified
1608 address (no domain) and a large number of spaces between the end of the
1609 name and the colon was received, the reception process suffered a buffer
1610 overflow, and (when I tested it) crashed. This was caused by some obsolete
1611 code that should have been removed. The fix is to remove it!
1613 3. When running in the test harness, delay a bit after writing a bounce
1614 message to get a bit more predictability in the log output.
1616 4. Added a call to search_tidyup() just before forking a reception process. In
1617 theory, someone could use a lookup in the expansion of smtp_accept_max_
1618 per_host which, without the tidyup, could leave open a database connection.
1620 5. Added the variables $recipient_data and $sender_data which get set from a
1621 lookup success in an ACL "recipients" or "senders" condition, or a router
1622 "senders" option, similar to $domain_data and $local_part_data.
1624 6. Moved the writing of debug_print from before to after the "senders" test
1627 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1628 problems for message scanning, either using a data ACL, or using
1629 local_scan() because the Received: header was not generated till after they
1630 were called (in order to set the time as the time of reception completion).
1631 I have revised the way this works. The header is now generated after the
1632 body is received, but before the ACL or local_scan() are called. After they
1633 are run, the timestamp in the header is updated.
1639 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1640 before starting a queue runner without re-exec. This happened only when
1641 deliver_drop_privilege was set or when the Exim user was set to root. The
1642 effect of the bug was that timeouts during subsequent deliveries caused
1643 crashes instead of being properly handled. The handler is now left at its
1644 default (and expected) setting.
1646 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1647 message, again when deliver_drop_privilege is set or Exim is run as root.
1648 The bug described in (1) was not present in this case, but the tidying up
1649 of the other signals was missing. I have made the two cases consistent.
1651 3. The ignore_target_hosts setting on a manualroute router was being ignored
1652 for hosts that were looked up using the /MX notation.
1654 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1657 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1658 operated on the sender address. After changing the $sender_address to <>
1659 for the sender address verify, Exim was re-instated it as the original
1660 (before rewriting) address, but remembering that it had rewritten it, so it
1661 wasn't rewriting it again. This bug also had the effect of breaking the
1662 sender address verification caching when the sender address was rewritten.
1664 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1665 This has been changed so that if the ip literal address matches
1666 ignore_target_hosts, the router declines.
1668 7. Added expansion conditions match_domain, match_address, and match_local_
1669 part (NOT match_host).
1671 8. The placeholder for the Received: header didn't have a length field set.
1673 9. Added code to Exim itself and to exim_lock to test for a specific race
1674 condition that could lead to file corruption when using MBX delivery. The
1675 issue is with the lockfile that is created in /tmp. If this file is removed
1676 after a process has opened it but before that process has acquired a lock,
1677 there is the potential for a second process to recreate the file and also
1678 acquire a lock. This could lead to two Exim processes writing to the file
1679 at the same time. The added code performs the same test as UW imapd; it
1680 checks after acquiring the lock that its file descriptor still refers to
1681 the same named file.
1683 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1684 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1687 11. Added the smtp_active_hostname option. If used, this will typically be made
1688 to depend on the incoming interface address. Because $interface_address is
1689 not set up until the daemon has forked a reception process, error responses
1690 that can happen earlier (such as "too many connections") no longer contain
1693 12. If an expansion in a condition on a "warn" statement fails because a lookup
1694 defers, the "warn" statement is abandoned, and the next ACL statement is
1695 processed. Previously this caused the whole ACL to be aborted.
1697 13. Added the iplsearch lookup type.
1699 14. Added ident_timeout as a log selector.
1701 15. Added tls_certificate_verified as a log selector.
1703 16. Added a global option tls_require_ciphers (compare the smtp transport
1704 option of the same name). This controls incoming TLS connections.
1706 17. I finally figured out how to make tls_require_ciphers do a similar thing
1707 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1708 before starting the TLS session.
1710 18. Tabs are now shown as \t in -bP output.
1712 19. If the log selector return_path_on_delivery was set, Exim crashed when
1713 bouncing a message because it had too many Received: header lines.
1715 20. If two routers both had headers_remove settings, and the first one included
1716 a superfluous trailing colon, the final name in the first list and the
1717 first name in the second list were incorrectly joined into one item (with a
1718 colon in the middle).
1724 1. Added -C and -D options to the exinext utility, mainly to make it easier
1725 to include in the automated testing, but these could be helpful when
1726 multiple configurations are in use.
1728 2. The exinext utility was not formatting the output nicely when there was
1729 an alternate port involved in the retry record key, nor when there was a
1730 message id as well (for retries that were specific to a specific message
1731 and a specific host). It was also confused by IPv6 addresses, because of
1732 the additional colons they contain. I have fixed the IPv4 problem, and
1733 patched it up to do a reasonable job for IPv6.
1735 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1736 delivery, the log line now contains "pipelined" if PIPELINING was used.
1738 4. An SMTP transport process used to panic and die if the bind() call to set
1739 an explicit outgoing interface failed. This has been changed; it is now
1740 treated in the same way as a connect() failure.
1742 5. A reference to $sender_host_name in the part of a conditional expansion
1743 that was being skipped was still causing a DNS lookup. This no longer
1746 6. The def: expansion condition was not recognizing references to header lines
1747 that used bh_ and bheader_.
1749 7. Added the _cache feature to named lists.
1751 8. The code for checking quota_filecount in the appendfile transport was
1752 allowing one more file than it should have been.
1754 9. For compatibility with Sendmail, the command line option
1762 and sets the incoming protocol and host name (for trusted callers). The
1763 host name and its colon can be omitted when only the protocol is to be set.
1764 Note the Exim already has two private options, -pd and -ps, that refer to
1765 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1766 "s", but I don't think that's a major issue.
1768 10. A number of refactoring changes to the code, none of which should affect
1771 (a) The number of logging options was getting close to filling up the
1772 32-bit word that was used as a bit map. I have split them into two classes:
1773 those that are passed in the argument to log_write(), and those that are
1774 only ever tested independently outside of that function. These are now in
1775 separate 32-bit words, so there is plenty of room for expansion again.
1776 There is no change in the user interface or the logging behaviour.
1778 (b) When building, for example, log lines, the code previously used a
1779 macro that called string_cat() twice, in order to add two strings. This is
1780 not really sufficiently general. Furthermore, there was one instance where
1781 it was actually wrong because one of the argument was used twice, and in
1782 one call a function was used. (As it happened, calling the function twice
1783 did not affect the overall behaviour.) The macro has been replaced by a
1784 function that can join an arbitrary number of extra strings onto a growing
1787 (c) The code for expansion conditions now uses a table and a binary chop
1788 instead of a serial search (which was left over from when there were very
1789 few conditions). Also, it now recognizes conditions like "pam" even when
1790 the relevant support is not compiled in: a suitably worded error message is
1791 given if an attempt is made to use such a condition.
1793 11. Added ${time_interval:xxxxx}.
1795 12. A bug was causing one of the ddress fields not to be passed back correctly
1796 from remote delivery subprocesses. The field in question was not being
1797 subsequently used, so this caused to problems in practice.
1799 13. Added new log selectors queue_time and deliver_time.
1801 14. Might have fixed a bug in maildirsizefile handling that threw up
1802 "unexpected character" debug warnings, and recalculated the data
1803 unnecessarily. In any case, I expanded the warning message to give more
1806 15. Added the message "Restricted characters in address" to the statements in
1807 the default ACL that block characters like @ and % in local parts.
1809 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1810 Three changes have been made:
1812 (a) There was a serious bug; a negative response to MAIL caused the whole
1813 recipient domain to be cached as invalid, thereby blocking all messages
1814 to all local parts at the same domain, from all senders. This bug has
1815 been fixed. The domain is no longer cached after a negative response to
1816 MAIL if the sender used is not empty.
1818 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1821 (c) A new callout option, "use_sender" has been added for people who want
1822 the modified behaviour.
1828 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1831 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1832 indeed breaks things for older releases.
1834 3. Added additional logging to the case where there is a problem reading data
1835 from a filter that is running in a subprocess using a pipe, in order to
1836 try to track down a specific problem.
1838 4. Testing facility fudge: when running in the test harness and attempting
1839 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1840 getting "No route to host". Convert this to a timeout.
1842 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1845 6. Some OS don't have socklen_t but use size_t instead. This affects the
1846 fifth argument of getsockopt() amongst other things. This is now
1847 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1848 can be set for individual OS. I have set it for SunOS5, OSF1, and
1849 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1850 some earlier ones do not.
1852 7. Change 4.30/15 was not doing the test caselessly.
1854 8. The standard form for an IPv6 address literal was being rejected by address
1855 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1856 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1857 this, as well as the form without the "IPv6" on the front (but only when
1858 address literals are enabled, of course).
1860 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1862 10. Exim crashed if a message with an empty sender address specified by -f
1863 encountered a router with an errors_to setting. This could be provoked only
1864 by a command such as
1868 where an empty string was supplied; "<>" did not hit this bug.
1870 11. Installed PCRE release 4.5.
1872 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1873 remained set. It is now erased.
1875 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1876 times from message ids (which are base 36 rather than the normal 62).
1878 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1879 were being counted as actual protocol errors, and logged if the log
1880 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1881 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1882 and DATA following a set of rejected RCPTs do not count as protocol errors.
1883 In other words, Exim assumes they were pipelined, though this may not
1884 actually be the case. Of course, in all cases the client gets an
1885 appropriate error code.
1887 15. If a lookup fails in an ACL condition, a message about the failure may
1888 be available; it is used if testing the ACL cannot continue, because most
1889 such messages specify what the cause of the deferral is. However, some
1890 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1891 that caused an old message to be retained and used if a later statement
1892 caused a defer, replacing the real cause of the deferral.
1894 16. If an IP address had so many PTR records that the DNS lookup buffer
1895 was not large enough to hold them, Exim could crash while trying to process
1896 the truncated data. It now detects and logs this case.
1898 17. Further to 4.21/58, another change has been made: if (and only if) the
1899 first line of a message (the first header line) ends with CRLF, a bare LF
1900 in a subsequent header line has a space inserted after it, so as not to
1901 terminate the header.
1903 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1904 unnecessarily, used atoi() instead of strtol(), and didn't check the
1905 termination when getting file sizes from file names by regex.
1907 19. Completely re-implemented the support for maildirsize files, in the light
1908 of a number of problems with the previous contributed implementation
1909 (4.30/29). In particular:
1911 . If the quota is zero, the maildirsize file is maintained, but no quota is
1914 . If the maildir directory does not exist, it is created before any attempt
1915 to write a maildirsize file.
1917 . The quota value in the file is just a cache; if the quota is changed in
1918 the transport, the new value overrides.
1920 . A regular expression is available for excluding directories from the
1923 20. The autoreply transport checks the characters in options that define the
1924 message's headers; it allows continued headers, but it was checking with
1925 isspace() after an embedded newline instead of explicitly looking for a
1928 21. If all the "regular" hosts to which an address was routed had passed their
1929 expiry times, and had not reached their retry times, the address was
1930 bounced, even if fallback hosts were defined. Now Exim should go on to try
1933 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1934 equivalent code in the SMTP transport. Some hosts send humungous responses
1935 to HELO/EHLO, more than 1024 it seems.
1937 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1938 gives compiler warnings in some environments. I've now done it "properly",
1941 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1942 (because of problems with the built-in one) was declared to return uschar *
1943 instead of char *, causing compiler failure.
1945 25. Fixed a file descriptor leak when processing alias/forward files.
1947 26. Fixed a minor format string issue in dbfn.c.
1949 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1951 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1952 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1955 29. When a sender address is verified, it is cached, to save repeating the test
1956 when there is more than one recipient in a message. However, when the
1957 verification involves a callout, it is possible for different callout
1958 options to be set for different recipients. It is too complicated to keep
1959 track of this in the cache, so now Exim always runs a verification when a
1960 callout is required, relying on the callout cache for the optimization.
1961 The overhead is duplication of the address routing, but this should not be
1964 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
1965 to be verified with callout=postmaster, and the main callout worked but the
1966 postmaster check failed, the verification correctly failed. However, if a
1967 subsequent RCPT command asked for sender verification *without* the
1968 postmaster check, incorrect caching caused this verification also to fail,
1971 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1972 it was not caching the DNS options (qualify_single, search_parents) that
1973 were used when the lookup failed. A subsequent lookup with different
1974 options therefore always gave the same answer, though there were cases
1975 where it should not have. (Example: a "domains = !$mx_any" option on a
1976 dnslookup router: the "domains" option is always processed without any
1977 widening, but the router might have qualify_single set.) Now Exim uses the
1978 cached value only when the same options are set.
1980 32. Added John Jetmore's "exipick" utility to the distribution.
1982 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
1983 than a timeout (e.g. a certificate is required, and is not provided), an
1984 Exim server now closes the connection immediately. Previously it waited for
1985 the client to close - but if the client is SSL, it seems that they each
1986 wait for each other, leading to a delay before one of them times out.
1988 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1989 maintained 0.8.x compatibility because I don't think many are using it, and
1990 it is clearly obsolete.
1992 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1995 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1996 client certificate was expired. A simple patch fixes this, though I don't
1997 understand the full logic of why the verify callback is called multiple
2000 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2001 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2002 which causes problems with some clients (such as the Certicom SSL Plus
2003 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2004 disables the coutermeasure allowing Eudora to connect."
2006 38. Exim was not checking that a write() to a log file succeeded. This could
2007 lead to Bad Things if a log got too big, in particular if it hit a file
2008 size limit. Exim now panics and dies if it cannot write to a log file, just
2009 as it does if it cannot open a log file.
2011 39. Modified OS/Makefile-Linux so that it now contains
2013 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2015 The two -D definitions ensure that Exim is compiled with large file
2016 support, which makes it possible to handle log files that are bigger than
2019 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2020 instance) a domain was checked against a named list that involved a lookup,
2021 causing $domain_data to be set, then another domain was checked against the
2022 same list, then the first domain was re-checked, the value of $domain_data
2023 after the final check could be wrong. In particular, if the second check
2024 failed, it could be set empty. This bug probably also applied to
2027 41. The strip_trailing_dot option was not being applied to the address given
2028 with the -f command-line option.
2030 42. The code for reading a message's header from the spool was incrementing
2031 $received_count, but never initializing it. This meant that the value was
2032 incorrect (doubled) while delivering a message in the same process in which
2033 it was received. In the most common configuration of Exim, this never
2034 happens - a fresh exec is done - but it can happen when
2035 deliver_drop_privilege is set.
2037 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2038 it now includes up to 150 characters of the unexpected data in the log
2041 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2042 and building data strings. The size of both of these buffers was 10 000
2043 bytes - far larger than anybody would *ever* want, thought I. Needless to
2044 say, somebody hit the limit. I have increased the maximum line length to
2045 20 000 and the maximum data length of concatenated lines to 100 000. I have
2046 also fixed two bugs, because there was no checking on these buffers. Tsk,
2047 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2048 buffer is too small.
2050 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2051 lsearch lookups. Now it does.
2053 46. When parsing a route_list item in a manualroute router, a fixed-length
2054 buffer was used for the list of hosts. I made this 1024 bytes long,
2055 thinking that nobody would ever have a list of hosts that long. Wrong.
2056 Somebody had a whole pile of complicated expansion conditions, and the
2057 string was silently truncated, leading to an expansion error. It turns out
2058 that it is easier to change to an unlimited length (owing to other changes
2059 that have happened since this code was originally written) than to build
2060 structure for giving a limitation error. The length of the item that
2061 expands into the list of hosts is now unlimited.
2063 47. The lsearch lookup could not handle data where the length of text line was
2064 more than 4095 characters. Such lines were truncated, leading to shortened
2065 data being returned. It should now handle lines of any length.
2067 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2068 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2071 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2072 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2073 no longer happens in comment lines. A list of replacements is now placed
2074 at the head of all of the source files, except those whose only change is
2075 to replace PERL_COMMAND in the very first #! line.
2077 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2078 messages on the queue, with a bottom-up merge sort, using code contributed
2079 by Michael Haardt. This should make operations like -bp somewhat faster on
2080 large queues. It won't affect queue runners, except when queue_run_in_order
2083 51. Installed eximstats 1.31 in the distribution.
2085 52. Added support for SRV lookups to the dnslookup router.
2087 53. If an ACL referred to $message_body or $message_body_end, the value was not
2088 reset for any messages that followed in the same SMTP session.
2090 54. The store-handling optimization for building very long strings was not
2091 differentiating between the different store pools. I don't think this
2092 actually made any difference in practice, but I've tidied it.
2094 55. While running the routers to verify a sender address, $sender_address
2095 was still set to the sender address. This is wrong, because when routing to
2096 send a bounce to the sender, it would be empty. Therefore, I have changed
2097 it so that, while verifying a sender address, $sender_address is set to <>.
2098 (There is no change to what happens when verifying a recipient address.)
2100 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2101 target A or AAAA records (if not already returned) without resetting the
2102 qualify_single or search_parents options of the DNS resolver. These are
2103 inappropriate in this case because the targets of MX and SRV records must
2104 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2105 target that, when qualified, matched something in the local domain. These
2106 two options are now turned off when doing these lookups.
2108 57. It seems that at least some releases of Reiserfs (which does not have the
2109 concept of a fixed number of inodes) returns zero and not -1 for the
2110 number of available inodes. This interacted badly with check_spool_inodes,
2111 which assumed that -1 was the "no such thing" setting. What I have done is
2112 to check that the total number of inodes is greater than zero before doing
2113 the test of how many are available.
2115 58. When a "warn" ACL statement has a log_message modifier, the message is
2116 remembered, and not repeated. This is to avoid a lot of repetition when a
2117 message has many recipients that cause the same warning to be written.
2118 Howewer, Exim was preserving the list of already written lines for an
2119 entire SMTP session, which doesn't seem right. The memory is now reset if a
2120 new message is started.
2122 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2123 debugging output unless log_rewrite was also set.
2125 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2126 of (int)(handle) when we know that handle contains (void *)(-1).
2128 61. The Exim daemon panic-logs an error return when it closes the incoming
2129 connection. However "connection reset by peer" seems to be common, and
2130 isn't really an error worthy of noting specially, so that particular error
2133 62. When Exim is trying to find all the local interfaces, it used to panic and
2134 die if the ioctl to get the interface flags failed. However, it seems that
2135 on at least one OS (Solaris 9) it is possible to have an interface that is
2136 included in the list of interfaces, but for which you get a failure error
2137 for this call. This happens when the interface is not "plumbed" into a
2138 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2139 failure of the "get flags" call assumes that the interface is down.
2141 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2142 makes life easier for people who are doing arithmetic on fields extracted
2143 from dates, where you often get leading zeros that should not be
2144 interpreted as octal.
2146 64. Added qualify_domain to the redirect router, to override the global
2149 65. If a pathologically long header line contained very many addresses (the
2150 report of this problem mentioned 10 000) and each of them was rewritten,
2151 Exim could use up a very large amount of memory. (It kept on making new
2152 copies of the header line as it rewrote, and never released the old ones.)
2153 At the expense of a bit more processing, the header rewriting function has
2154 been changed so that it no longer eats memory in this way.
2156 66. The generation of the Received: header has been moved from the time that a
2157 message starts to be received, to the time that it finishes. The timestamp
2158 in the Received: header should now be very close to that of the <= log
2159 line. There are two side-effects of this change:
2161 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2162 logged header lines no longer include the local Received: line, because
2163 it has not yet been created. The same applies to a copy of the message
2164 that is returned to a non-SMTP sender when a message is rejected.
2166 (b) When a filter file is tested using -bf, no additional Received: header
2167 is added to the test message. After some thought, I decided that this
2170 This change does not affect the value of $received_for. It is still set
2171 after address rewriting, but before local_scan() is called.
2173 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2175 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2176 gave an unhelpful panic error message, and a defer error. I have managed to
2177 change this behaviour so that it now rejects any supplied certificate,
2178 which seems right, as the list of acceptable certificates is empty.
2180 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2181 gave an unhelpful defer error. I have not managed to make this reject any
2182 supplied certificates, but the error message it gives is "no certificate
2183 supplied", which is not helpful.
2185 70. exigrep's output now also includes lines that are not associated with any
2186 message, but which match the given pattern. Implemented by a patch from
2187 Martin Sluka, which also tidied up the Perl a bit.
2189 71. Recipient callout verification, like sender verification, was using <> in
2190 the MAIL FROM command. This isn't really the right thing, since the actual
2191 sender may affect whether the remote host accepts the recipient or not. I
2192 have changed it to use the actual sender in the callout; this means that
2193 the cache record is now keyed on a recipient/sender pair, not just the
2194 recipient address. There doesn't seem to be a real danger of callout loops,
2195 since a callout by the remote host to check the sender would use <>.
2196 [SEE ABOVE: changed after hitting problems.]
2198 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2199 temporary errors. However, in the case of such a code being given after
2200 the end of a data transmission (i.e. after ".") Exim was failing to write
2201 a retry record for the message. (Yes, there was some broken host that was
2202 actually sending 8xx at this point.)
2204 73. An unknown lookup type in a host list could cause Exim to panic-die when
2205 the list was checked. (An example that provoked this was putting <; in the
2206 middle of a list instead of at the start.) If this happened during a DATA
2207 ACL check, a -D file could be left lying around. This kind of configuration
2208 error no longer causes Exim to die; instead it causes a defer errror. The
2209 incident is still logged to the main and panic logs.
2211 74. Buglet left over from Exim 3 conversion. The message "too many messages
2212 in one connection" was written to the rejectlog but not the mainlog, except
2213 when address rewriting (yes!) was being logged.
2215 75. Added write_rejectlog option.
2217 76. When a system filter was run not as root (that is, when system_filter_user
2218 was set), the values of the $n variables were not being returned to the
2219 main process; thus, they were not subsequently available in the $sn
2222 77. Added +return_path_on_delivery log selector.
2224 78. A connection timeout was being treated differently from recipients deferred
2225 when testing hosts_max_try with a message that was older than the host's
2226 retry timeout. (The host should not be counted, thus allowing all hosts to
2227 be tried at least once before bouncing.) This may have been the cause of an
2228 occasionally reported bug whereby a message would remain on the queue
2229 longer than the retry timeout, but would be bounced if a delivery was
2230 forced. I say "may" because I never totally pinned down the problem;
2231 setting up timeout/retry tests is difficult. See also the next item.
2233 79. The ultimate address timeout was not being applied to errors that involved
2234 a combination of host plus message (for example, a timeout on a MAIL
2235 command). When an address resolved to a number of possible hosts, and they
2236 were not all tried for each delivery (e.g. because of hosts_max_try), a
2237 message could remain on the queue longer than the retry timeout.
2239 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2242 81. Fixed an obscure SMTP outgoing bug which required at least the following
2243 conditions: (a) there was another message waiting for the same server;
2244 (b) the server returned 5xx to all RCPT commands in the first message so
2245 that the message was not completed; (c) the server dropped the connection
2246 or gave a negative response to the RSET that Exim sends to abort the
2247 transaction. The observed case was a dropped connection after DATA that had
2248 been sent in pipelining mode. That is, the server had advertised PIPELINING
2249 but was not implementing it correctly. The effect of the bug was incorrect
2250 behaviour, such as trying another host, and this could lead to a crash.
2256 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2257 and daemon.c were passed as pointers to ints; they should have been
2258 pointers to socklen_t variables (which are typically unsigned ints).
2260 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2263 3. Fixed a really odd bug that affected only the testing scheme; patching a
2264 certain fixed string in the binary changed the value of another string that
2265 happened to be identical to the end of the original first string.
2267 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2268 name", it returns that address as the IP address. On some operating
2269 systems (e.g. Solaris), it also passes back the IP address string as the
2270 "host name". However, on others (e.g. Linux), it passes back an empty
2271 string. Exim wasn't checking for this, and was changing the host name to an
2272 empty string, assuming it had been canonicized.
2274 5. Although rare, it is permitted to have more than one PTR record for a given
2275 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2276 all the names associated with an address, because they do in Solaris.
2277 However, it seems that they do not in Linux for data that comes from the
2278 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2279 I found this out when I moved to a new Linux workstation and tried to run
2280 the Exim test suite.
2282 To get round this problem I have changed the code so that it now does its
2283 own call to the DNS to look up PTR records when searching for a host name.
2284 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2285 addresses that are only in /etc/hosts are still found.
2287 This behaviour is, however, controlled by an option called host_lookup_
2288 order, which defaults to "bydns:byaddr". If people want to use the other
2289 order, or indeed, just use one or the other means of lookup, they can
2290 specify it in this variable.
2292 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2293 some operating systems, this comes back from gethostbyaddr() as an empty
2294 string, and this is what Exim used to test for. However, it seems that in
2295 other systems, "." is yielded. Exim now tests for this case too.
2297 7. The values of check_spool_space and check_log_space are now held internally
2298 as a number of kilobytes instead of an absolute number of bytes. If a
2299 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2300 kilobyte. This means that much larger values can be stored.
2302 8. Exim monitor: an attempt to get the action menu when not actually pointing
2303 at a message produces an empty menu entitled "No message selected". This
2304 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2305 no entries in it ("Shell widget menu has zero width and/or height"). So I
2306 have added a single, blank menu entry in this case.
2308 9. Added ${quote_local_part.
2310 10. MIME decoding is now applied to the contents of Subject: header lines when
2313 11. Now that a reference to $sender_host_address automatically causes a reverse
2314 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2315 host lookup before query-style lookups in lists that might use this
2316 variable. This has therefore been abolished, and the "net-" prefix is no
2317 longer necessary for query-style lookups.
2319 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2320 have been a typo for LFLAGS, so it has been changed.
2322 13. The install script calls Exim with "-C /dev/null" in order to find the
2323 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2324 to be output. Howeve, since Exim outputs its version number before the
2325 error, it didn't break the script. It just looked ugly. I fixed this by
2326 always allowing "-C /dev/null" if the caller is root.
2328 14. Ignore overlarge ACL variable number when reading spool file - insurance
2329 against a later release with more variables having written the file.
2331 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2332 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2333 this, as well as the form without the "IPv6" on the front.
2335 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2336 OS/Makefile-Darwin file.
2338 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2339 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2341 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2342 variable. However, some OS use ints for inodes. I've added cast to long int
2343 to get rid of the compiler warning.
2345 19. I had forgotten to lock out "/../" in configuration file names when
2346 ALT_CONFIG_PREFIX was set.
2348 20. Routers used for verification do not need to specify transports. However,
2349 if such a router generated a host list, and callout was configured, Exim
2350 crashed, because it could not find a port number from the (non-existent)
2351 transport. It now assumes port 25 in this circumstance.
2353 21. Added the -t option to exigrep.
2355 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2356 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2357 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2358 them is set, LOOKUP_LSEARCH is forced.
2360 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2362 24. Added sender and host information to the "rejected by local_scan()" log
2363 line; previously there was no indication of these.
2365 25. Added .include_if_exists.
2367 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2368 Linux. It turns out that not all file systems support this. Apparently some
2369 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2370 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2371 not supported on the file descriptor, is now ignored when Exim is trying to
2372 sync a directory. This applies only to Linux.
2374 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2376 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2377 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2380 29. Added support for maildirsize files from supplied patch (modified a bit).
2382 30. The use of :fail: followed by an empty string could lead Exim to respond to
2383 sender verification failures with (e.g.):
2385 550 Verification failed for <xxx>
2386 550 Sender verify failed
2388 where the first response line was missing the '-' that indicates it is not
2389 the final line of the response.
2391 31. The loop for finding the name of the user that called Exim had a hardwired
2392 limit of 10; it now uses the value of finduser_retries, which is used for
2393 all other user lookups.
2395 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2398 33. Exim was neglecting to zero errno before one call of strtol() when
2399 expanding a string and expecting an integer value. On some systems this
2400 resulted in spurious "integer overflow" errors. Also, it was casting the
2401 result into an int without checking.
2403 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2404 did not work. The code looks as if it has *never* worked, though it appears
2405 to have been documented since at least releast 1.62. I have made it work.
2407 35. The "timeout_DNS" error in retry rules, also documented since at least
2408 1.62, also never worked. As it isn't clear exactly what this means, and
2409 clearly it isn't a major issue, I have abolished the feature by treating it
2410 as "timeout", and writing a warning to the main and panic logs.
2412 36. The display of retry rules for -brt wasn't always showing the error code
2415 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2416 timeout_connect_A, timeout_connect_MX.
2418 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2419 to the empty sender.
2421 39. The daemon was not analysing the content of -oX till after it had closed
2422 stderr and disconnected from the controlling terminal. This meant that any
2423 syntax errors were only noted on the panic log, and the return code from
2424 the command was 0. By re-arranging the code a little, I've made the
2425 decoding happen first, so such errors now appear on stderr, and the return
2426 code is 1. However, the actual setting up of the sockets still happens in
2427 the disconnected process, so errors there are still only recorded on the
2430 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2431 connections (as happens on some IP stacks) was logged at start up time as
2432 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2433 it from "IPv6 and IPv4", which means that two separate sockets are being
2436 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2437 says whether AF_INET or AF_INET6 was passed as an argument.
2439 42. Exiwhat output was messed up when time zones were included in log
2442 43. Exiwhat now gives more information about the daemon's listening ports,
2443 and whether -tls-on-connect was used.
2445 44. The "port" option of the smtp transport is now expanded.
2447 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2448 silently ignored. Now an error message is written to the main and panic
2451 46. There's a new ACL modifier called "logwrite" which writes to a log file
2452 as soon as it is encountered.
2454 47. Added $local_user_uid and $local_user_gid at routing time.
2456 48. Exim crashed when trying to verify a sender address that was being
2459 49. Exim was recognizing only a space character after ".include". It now also
2460 recognizes a tab character.
2462 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2463 extracting the relevant information from the specification. The man page no
2464 longer contains scrambled data for the -d option, and I've added a section
2465 at the front about calling Exim under different names.
2467 51. Added "extra_headers" argument to the "mail" command in filter files.
2469 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2472 53. Installed eximstats 1.29.
2474 54. Added transport_filter_timeout as a generic transport option.
2476 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2477 Cc: header lines. This was required by RFC 822, but it not required by RFC
2480 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2481 incoming messages that did not have them. Now it does so only if the
2482 message originates locally, that is, if there is no associated remote host
2483 address. When Resent- header lines are present, this applies to the Resent-
2484 lines rather than the non-Resent- lines.
2486 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2487 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2489 58. Messages for configuration errors now include the name of the main
2490 configuration file - useful now that there may be more than one file in a
2491 list (.included file names were always shown).
2493 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2494 for those rare installations that do not start the daemon as root or run it
2495 setuid root. I've cut out the call to initgroups() if the daemon is not
2498 60. The Exim user and group can now be bound into the binary as text strings
2499 that are looked up at the start of Exim's processing.
2501 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2503 62. Added $mailstore_basename variable.
2505 63. Installed patch to sieve.c from Michael Haardt.
2507 64. When Exim failed to open the panic log after failing to open the main log,
2508 the original message it was trying to log was written to stderr and debug
2509 output, but if they were not available (the usual case in production), it
2510 was lost. Now it is written to syslog before the two lines that record the
2511 failures to open the logs.
2513 65. Users' Exim filters run in subprocesses under the user's uid. It is
2514 possible for a "deliver" command or an alias in a "personal" command to
2515 provoke an address rewrite. If logging of address rewriting is configured,
2516 this fails because the process is not running as root or exim. There may be
2517 a better way of dealing with this, but for the moment (because 4.30 needs
2518 to be released), I have disabled address rewrite logging when running a
2519 filter in a non-root, non-exim process.
2525 1. The buildconfig auxiliary program wasn't quoting the value set for
2526 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2527 not defined. This bug was masked in 4.22 by the effect that was fixed in
2530 2. Some messages that were rejected after a message id was allocated were
2531 shown as "incomplete" by exigrep. It no longer does this for messages that
2532 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2534 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2535 have allow_domain_literals set, the ID did not get logged in the <= line.
2536 Domain literals are now always recognized in Message-ID: header lines.
2538 4. The first argument for a ${extract expansion item is the key name or field
2539 number. Leading and trailing spaces in this item were not being ignored,
2540 causing some misleading effects.
2542 5. When deliver_drop_privilege was set, single queue runner processes started
2543 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2544 same command in the process it spins off) were not dropping privilege.
2546 6. When the daemon running as "exim" started a queue runner, it always
2547 re-executed Exim in the spun-off process. This is a waste of effort when
2548 deliver_drop_privilege is set. The new process now just calls the
2549 queue-runner function directly.
2555 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2558 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2559 ignored. Though the use of -oP was forcing the writing of a pid file, it
2560 was always written to the default place.
2562 3. If the message "no IP address found for host xxxx" is generated during
2563 incoming verification, it is now followed by identification of the incoming
2564 connection (so you can more easily find what provoked it).
2566 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2568 5. Added some features to "harden" Exim a bit more against certain attacks:
2570 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2571 be put in Local/Makefile. This is like the never_users runtime option,
2572 but it cannot be overridden. The default setting is "root".
2574 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2575 prefix string with which any file named in a -C command line option
2578 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2579 is retained for -C and -D only if the caller of Exim is root. Without
2580 it, the exim user may also use -C and -D and retain privilege.
2582 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2583 command line option is disabled.
2585 6. Macro names set by the -D option must start with an upper case letter, just
2586 like macro names defined in the configuration file.
2588 7. Added "dereference=" facility to LDAP.
2590 8. Two instances of the typo "uknown" in the source files are fixed.
2592 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2593 the Configure-Makefile script screwed up while processing it.
2595 10. Incorporated PCRE 4.4.
2597 11. The SMTP synchronization check was not operating right at the start of an
2598 SMTP session. For example, it could not catch a HELO sent before the client
2599 waited for the greeting. There is now a check for outstanding input at the
2600 point when the greeting is written. Because of the duplex, asynchronous
2601 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2602 way, but not yet received, when the check is performed.
2604 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2605 on TCP/IP sockets, because this apparently causes some broken clients to
2608 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2609 unchanged) from the Cygwin maintainer.
2611 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2612 was supported instead of testing for mbx. Effectively a typo.
2614 15. The spa authenticator server code was not checking that the input it
2615 received was valid base64.
2617 16. The debug output line for the "set" modifier in ACLs was not showing the
2618 name of the variable that was being set.
2620 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2622 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2623 lookup on its own. Something else was needed to trigger the lookup. For
2624 example, a match in host_lookup or the need for a host name in a host list.
2625 Now, if $sender_host_name is referenced and the host name has not yet been
2626 looked up, a lookup is performed. If the lookup fails, the variable remains
2627 empty, and $host_lookup_failed is set to "1".
2629 19. Added "eqi" as a case-independent comparison operator.
2631 20. The saslauthd authentication condition could segfault if neither service
2632 nor realm was specified.
2634 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2635 error message that was logged was misleading, and incoming SMTP
2636 connections were dropped. The message is now more accurate, and temporary
2637 errors are given to SMTP connections.
2639 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2640 (except RSET) with a 421 error, until QUIT is received. However, it was
2641 failing to send a response to QUIT.
2643 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2644 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2645 ACL is now run after the helo_try_verify_hosts code.
2647 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2648 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2649 case-independent, so other case variants are also recognized). Apparently
2650 some systems use these upper case variants.
2652 25. If more than two messages were waiting for the same host, and a transport
2653 filter was specified for the transport, Exim sent two messages over the
2654 same TCP/IP connection, and then failed with "socket operation on non-
2655 socket" when it tried to send the third.
2657 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2659 27. The extern definition of crypt16() in expand.c was not being excluded when
2660 the OS had its own crypt16() function.
2662 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2663 as a preferred synonym for return_size_limit, both as an option and as an
2666 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2668 30. Changed the default configuration ACL to relax the local part checking rule
2669 for addresses that are not in any local domains. For these addresses,
2670 slashes and pipe symbols are allowed within local parts, but the sequence
2671 /../ is explicitly forbidden.
2673 31. SPA server authentication was not clearing the challenge buffer before
2676 32. log_message in a "warn" ACL statement was writing to the reject log as
2677 well as to the main log, which contradicts the documentation and doesn't
2678 seem right (because no rejection is happening). So I have stopped it.
2680 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2681 However, I am unable to do any testing of this.
2683 34. Fixed an infelicity in the appendfile transport. When checking directories
2684 for a mailbox, to see if any needed to be created, it was accidentally
2685 using path names with one or more superfluous leading slashes; tracing
2686 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2688 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2689 contents are added to the log line that is written for every discarded
2690 recipient. (Previously a log_message setting was ignored.)
2692 36. The ${quote: operator now quotes the string if it is empty.
2694 37. The install script runs exim in order to find its version number. If for
2695 some reason other than non-existence or emptiness, which it checks, it
2696 could not run './exim', it was installing it with an empty version number,
2697 i.e. as "exim-". This error state is now caught, and the installation is
2700 38. An argument was missing from the function that creates an error message
2701 when Exim fails to connect to the socket for saslauthd authentication.
2702 This could cause Exim to crash, or give a corrupted message.
2704 39. Added isip, isip4, and isip6 to ${if conditions.
2706 40. The ACL variables $acl_xx are now saved with the message, and can be
2707 accessed later in routers, transports, and filters.
2709 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2710 strings in the file are not string-expanded.
2712 42. If a MAIL command specified a SIZE value that was too large to fit into an
2713 int variable, the check against message_size_limit failed. Such values are
2714 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2715 day this will have to be increased, but I don't think I want to be around
2716 when emails are that large.
2723 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2724 iconv() is not standard in FreeBSD.
2726 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2727 IPv6 enabled. The observed symptom was a segmentation fault on return from
2728 the function os_common_find_running_interfaces() in src/os.c.
2730 3. In the check_special_case() function in daemon.c I had used "errno" as an
2731 argument name, which causes warnings on some systems. This was basically a
2732 typo, since it was named "eno" in the comments!
2734 4. The code that waits for the clock to tick (at a resolution of some fraction
2735 of a second) so as to ensure message-id uniqueness was always waiting for
2736 at least one whole tick, when it could have waited for less. [This is
2737 almost certainly not relevant at current processor speeds, where it is
2738 unlikely to ever wait at all. But we try to future-proof.]
2740 5. The function that sleeps for a time interval that includes fractions of a
2741 second contained a race. It did not block SIGALRM between setting the
2742 timer, and suspending (a couple of lines later). If the interval was short
2743 and the sigsuspend() was delayed until after it had expired, the suspension
2744 never ended. On busy systems this could lead to processes getting stuck for
2747 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2748 process, before it forks any delivery processes. The open lookup caching
2749 mechanism meant that the open file or database connection was passed into
2750 the delivery process. The problem was that delivery processes always tidy
2751 up cached lookup data. This could cause a problem for the next delivery
2752 process started by the queue runner, because the external queue runner
2753 process does not know about the closure. So the next delivery process
2754 still has data in the lookup cache. In the case of a file lookup, there was
2755 no problem because closing a file descriptor in a subprocess doesn't affect
2756 the parent. However, if the lookup was caching a connection to a database,
2757 the connection was closed, and the second delivery process was likely to
2758 see errors such as "PGSQL: query failed: server closed the connection
2759 unexpectedly". The problem has been fixed by closing all cached lookups
2760 in a queue runner before running a delivery process.
2762 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2763 seem to have the "const" qualifier which it has on other OS. I've
2766 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2767 *of the same type* (client or server) with the same public name that an
2768 error should be diagnosed.
2770 9. When Exim looked up a host name for an IP address, but failed to find the
2771 original IP address when looking up the host name (a safety check), it
2772 output the message "<ip address> does not match any IP for NULL", which was
2773 confusing, to say the least. The bug was that the host name should have
2774 appeared instead of "NULL".
2776 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2777 user that is built into the binary, and the -C or -D options is used, root
2778 privilege is dropped before the configuration file is read. In addition,
2779 logging is switched to stderr instead of the normal log files. If the
2780 configuration then re-defines the Exim user, the unprivileged environment
2781 is probably not what is expected, so Exim logs a panic warning message (but
2784 However, if deliver_drop_privilege is set, the unprivileged state may well
2785 be exactly what is intended, so the warning has been cut out in that case,
2786 and Exim is allowed to try to write to its normal log files.
2792 1. smtp_return_error_details was not giving details for temporary sender
2793 or receiver verification errors.
2795 2. Diagnose a configuration error if two authenticators have the same public
2798 3. Exim used not to create the message log file for a message until the first
2799 delivery attempt. This could be confusing when incoming messages were held
2800 for policy or load reasons. The message log file is now created at the time
2801 the message is received, and an initial "Received" line is written to it.
2803 4. The automatically generated man page for command line options had a minor
2804 bug that caused no ill effects; however, a more serious problem was that
2805 the procedure for building the man page automatically didn't always
2806 operate. Consequently, release 4.20 contains an out-of-date version. This
2807 shouldn't happen again.
2809 5. When building Exim with embedded Perl support, the script that builds the
2810 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2811 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2813 6. The freeze_tell option was not being used for messages that were frozen on
2814 arrival, either by an ACL or by local_scan().
2816 7. Added the smtp_incomplete_transaction log selector.
2818 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2819 was accepting AUTH without a new EHLO.
2821 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2822 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2825 10. Logging of TCP/IP connections (when configured) now happens in the main
2826 daemon process instead of the child process, so that the TCP/IP connection
2827 count is more accurate (but it can never be perfect).
2829 11. The use of "drop" in a nested ACL was not being handled correctly in the
2830 outer ACL. Now, if condition failure induced by the nested "drop" causes
2831 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2832 or "require"), the connection is dropped.
2834 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2835 that yield "discard" can now be used with an "accept" or a "discard" verb,
2836 but an error is generated for any others (because I can't see a useful way
2837 to define what should happen).
2839 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2840 are now processed in the same way as lines in the Exim configuration file.
2841 In particular, continuation lines are supported.
2843 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2845 15. Added -ti meaning -t -i.
2847 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2848 domains, and warn by logging if others are found.
2850 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2851 returned from GIFCONF when Exim is trying to find the list of interfaces on
2852 a host. The code in os.c has been modified to copy each ifreq to an aligned
2853 structure in all cases.
2855 Also, in some cases, the returned ifreq's were being copied to a 'struct
2856 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2857 means the last couple of bytes of an IPv6 address could be chopped if the
2858 ifreq contained only a normal sockaddr (14 bytes storage).
2860 18. Named domain lists were not supported in the hosts_treat_as_local option.
2861 An entry such as +xxxx was not recognized, and was treated as a literal
2864 19. Ensure that header lines added by a DATA ACL are included in the reject log
2865 if the ACL subsequently rejects the message.
2867 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2868 MD5 (which is deprecated).
2870 21. When testing a filter file using -bf, Exim was writing a message when it
2871 took the sender from a "From " line in the message, but it was not doing so
2872 when it took $return_path from a Return-Path: header line. It now does.
2874 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2875 with a valid header line field name (a series of printing characters
2876 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2878 23. Changed "disc" in the source to "disk" to conform to the documentation and
2879 the book and for uniformity.
2881 24. Ignore Sendmail's -Ooption=value command line item.
2883 25. When execve() failed while trying to run a command in a pipe transport,
2884 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2885 could be confused with a return value of 69 from the command itself. This
2886 has been changed to 127, the value the shell returns if it is asked to run
2887 a non-existent command. The wording for the related log line suggests a
2888 non-existent command as the problem.
2890 26. If received_header_text expands to an empty string, do not add a Received:
2891 header line to the message. (Well, it adds a token one on the spool, but
2892 marks it "old" so that it doesn't get used or transmitted.)
2894 27. Installed eximstats 1.28 (addition of -nt option).
2896 28. There was no check for failure on the call to getsockname() in the daemon
2897 code. This can fail if there is a shortage of resources on the system, with
2898 ENOMEM, for example. A temporary error is now given on failure.
2900 29. Contrary to the C standard, it seems that in some environments, the
2901 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2902 program. Exim now does this explicitly; it affects the formatting of
2903 timestamps using strftime().
2905 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2906 complaints. I've now initialized all the variables, to avoid this.
2908 32. Header lines added by a system filter were not being "seen" during
2909 transport-time rewrites.
2911 33. The info_callback() function passed to OpenSSL is set up with type void
2912 (*)(SSL *, int, int), as described somewhere. However, when calling the
2913 function (actually a macro) that sets it up, the type void(*)() is
2914 expected. I've put in a cast to prevent warnings from picky compilers.
2916 34. If a DNS black list lookup found a CNAME record, but there were no A
2917 records associated with the domain it pointed at, Exim crashed.
2919 35. If a DNS black list lookup returned more than one A record, Exim ignored
2920 all but the first. It now scans all returned addresses if a particular IP
2921 value is being sought. In this situation, the contents of the
2922 $dnslist_value variable are a list of all the addresses, separated by a
2925 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2926 to accept a host name and all its aliases if the forward lookup for any of
2927 them yielded the IP address of the incoming connection. Now it accepts only
2928 those names whose forward lookup yields the correct IP address. Any other
2929 names are discarded. This closes a loophole whereby a rogue DNS
2930 administrator could create reverse DNS records to break through a
2931 wildcarded host restriction in an ACL.
2933 37. If a user filter or a system filter that ran in a subprocess used any of
2934 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2935 the wrong values were passed to the pipe command ($thisaddress had the
2936 value of $0, $0 had the value of $1, etc). This bug was introduced by
2937 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2939 38. Improved the line breaking for long SMTP error messages from ACLs.
2940 Previously, if there was no break point between 40 and 75 characters, Exim
2941 left the rest of the message alone. Two changes have been made: (a) I've
2942 reduced the minimum length to 35 characters; (b) if it can't find a break
2943 point between 35 and 75 characters, it looks ahead and uses the first one
2944 that it finds. This may give the occasional overlong line, but at least the
2945 remaining text gets split now.
2947 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2948 file descriptors might be low, and that setting 1000 would always raise it.
2949 It turns out that in some environments, the limit is already over 1000 and
2950 that lowering it causes trouble. So now Exim takes care not to decrease it.
2952 40. When delivering a message, the value of $return_path is set to $sender_
2953 address at the start of routing (routers may change the value). By an
2954 oversight, this default was not being set up when an address was tested by
2955 -bt or -bv, which affected the outcome if any router or filter referred to
2958 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2959 log only when "message" or "log_message" are set. However, if one of the
2960 conditions was an address verification, or a call to a nested ACL, the
2961 messages generated by the underlying test were being passed through. This
2962 no longer happens. The underlying message is available in $acl_verify_
2963 message for both "message" and "log_message" expansions, so it can be
2964 passed through if needed.
2966 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2967 new expansion $bh_ to give the encoded byte string without charset
2968 translation. Translation happens only if iconv() is available; HAVE_ICONV
2969 indicates this at build time. HEADERS_CHARSET gives the charset to
2970 translate to; headers_charset can change it in the configuration, and
2971 "headers charset" can change it in an individual filter file.
2973 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
2974 that creates RFC 2047 encoded "words" labels them as that charset instead
2975 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2976 expansion operator; (ii) when Exim creates a From: line for a local
2977 message; (iii) when a header line is rewritten to include a "phrase" part.
2979 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2980 buggy, causing it to skip the first lines of messages whose message ID
2981 ended in 'D'. This would not have bitten before Exim release 4.14, because
2982 message IDs were unlikely to end in 'D' before then. The effect was to have
2983 incorrect size information for certain domains.
2985 45. #include "config.h" was missing at the start of the crypt16.c module. This
2986 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2989 46. If there was a timeout during a "random" callout check, Exim treated it as
2990 a failure of the random address, and carried on sending RSET and the real
2991 address. If the delay was just some slowness somewhere, the response to the
2992 original RCPT would be taken as a response to RSET and so on, causing
2993 mayhem of various kinds.
2995 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2996 when I implemented it. It didn't allow for the fact that some option values
2997 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2998 the right test for positive values.
3000 48. Domain names in DNS records are case-independent. Exim always looks them up
3001 in lower case. Some resolvers return domain names in exactly the case they
3002 appear in the zone file, that is, they may contain uppercase letters. Not
3003 all resolvers do this - some return always lower case. Exim was treating a
3004 change of case by a resolver as a change of domain, similar to a widening
3005 of a domain abbreviation. This triggered its re-routing code and so it was
3006 trying to route what was effectively the same domain again. This normally
3007 caused routing to fail (because the router wouldn't handle the domain
3008 twice). Now Exim checks for this case specially, and just changes the
3009 casing of the domain that it ultimately uses when it transmits the message
3012 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3015 50. If a filter generated a file delivery with a non-absolute name (possible if
3016 no home directory exists for the router), the forbid_file option was not
3019 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3020 the existing equality matching.
3022 52. Exim was using ints instead of ino_t variables in some places where it was
3023 dealing with inode numbers.
3025 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3026 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3027 called TMPDIR, and if it finds it is different, it changes its value.
3029 54. The smtp_printf() function is now made available to local_scan() so
3030 additional output lines can be written before returning. There is also an
3031 smtp_fflush() function to enable the detection of a dropped connection.
3032 The variables smtp_input and smtp_batched_input are exported to
3035 55. Changed the default runtime configuration: the message "Unknown user"
3036 has been removed from the ACL, and instead placed on the localuser router,
3037 using the cannot_route_message feature. This means that any verification
3038 failures that generate their own messages won't get overridden. Similarly,
3039 the "Unrouteable address" message that was in the ACL for unverifiable
3040 relay addresses has also been removed.
3042 56. Added hosts_avoid_esmtp to the smtp transport.
3044 57. The exicyclog script was not checking for the esoteric option
3045 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3046 will work only if exicyclog is run under the appropriate euid.
3048 58. Following a discussion on the list, the rules by which Exim recognises line
3049 endings on incoming messages have been changed. The -dropcr and drop_cr
3050 options are now no-ops, retained only for backwards compatibility. The
3051 following line terminators are recognized: LF CRLF CR. However, special
3052 processing applies to CR:
3054 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3055 nor a local message in the state where . is a terminator.
3057 (ii) If a bare CR is encountered in a header line, an extra space is added
3058 after the line terminator so as not to end the header. The reasoning
3059 behind this is that bare CRs in header lines are most likely either
3060 to be mistakes, or people trying to play silly games.
3062 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3063 was being incorrectly given as 18 bytes larger than it should have been.
3064 This is a VOB (very old bug).
3066 60. This may never have affected anything current, but just in case it has:
3067 When the local host is found other than at the start of a list of hosts,
3068 the local host, those with the same MX, and any that follow, are discarded.
3069 When the list in question was part of a longer list of hosts, the following
3070 hosts (not currently being processed) were also being discarded. This no
3071 longer happens. I'm not sure if this situation could ever has previously
3074 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3077 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3078 header. This is something that is recommended in a new Internet Draft, and
3079 is something that is documented as being done by Sendmail. There are two
3080 possible values. For messages generated by the autoreply transport, Exim
3083 Auto-Submitted: auto-replied
3085 whereas for all other generated messages (e.g. bounces) it adds
3087 Auto-Submitted: auto-generated
3089 63. The "personal" condition in filters now includes a test for the
3090 Auto-Submitted: header. If it contains the string "auto-" the message it
3091 not considered personal.
3093 64. Added rcpt_include_affixes as a generic transport option.
3095 65. Added queue_only_override (default true).
3097 66. Added the syslog_duplication option.
3099 67. If what should have been the first header line of a message consisted of
3100 a space followed by a colon, Exim was mis-interpreting it as a header line.
3101 It isn't of course - it is syntactically invalid and should therefore be
3102 treated as the start of the message body. The misbehaviour could have
3103 caused a number of strange effects, including loss of data in subsequent
3104 header lines, and spool format errors.
3106 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3107 client host had authenticated. This control can now be exercised by an ACL
3108 for more flexibility.
3110 69. By default, callouts do not happen when testing with -bh. There is now a
3111 variant, -bhc, which does actually run the callout code, including
3112 consulting and updating the callout cache.
3114 70. Added support for saslauthd authentication, courtesy of Alexander
3117 71. If statvfs() failed on the spool or log directories while checking their
3118 size for availability, Exim confusingly gave the error "space shortage".
3119 Furthermore, in debugging mode it crashed with a floating point exception.
3120 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3121 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3122 serious problem, Exim now writes to the main and panic logs when this
3123 happens, with details of the failure. It then refuses to accept the
3124 incoming message, giving the message "spool directory problem" or "log
3125 directory problem" with a 421 code for SMTP messages.
3127 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3128 0, 1, and 2 exist, because some OS complain for execs without them (see
3129 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3130 descriptors. However, the code omitted to check that the open succeeded,
3131 causing mysterious errors if for some reason the permissions on /dev/null
3132 got screwed. Now Exim writes a message to the main and panic logs, and
3133 bombs out if it can't open /dev/null.
3135 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3136 interact so that it is all more flexible. It is supposed to remain
3137 backwards compatible. Also added extra_local_interfaces.
3139 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3140 to bomb out with an assertion failure - to the client this appears as a
3141 connection drop. This problem occurs in the part of the code that was taken
3142 from the Samba project. Fortunately, the assertion is in a very simple
3143 function, so I have fixed this by reproducing the function inline in the
3144 one place where it is called, and arranging for authentication to fail
3145 instead of killing the process with assert().
3147 75. The SPA client code was not working when the server requested OEM rather
3148 than Unicode encoding.
3150 76. Added code to make require_files with a specific uid setting more usable in
3151 the case where statting the file as root fails - usually a non-root-mounted
3152 NFS file system. When this happens and the failure is EACCES, Exim now
3153 forks a subprocess and does the per-uid checking as the relevant uid.
3155 77. Added process_log_path.
3157 78. If log_file_path was not explicitly set, a setting of check_log_space or
3158 check_log_inodes was ignored.
3160 79. If a space check for the spool or log partitions fails, the incident is now
3161 logged. Of course, in the latter case the data may get lost...
3163 80. Added the %p formatting code to string_format() so that it can be used to
3164 print addresses in debug_print(). Adjusted all the address printing in the
3165 debugging in store.c to use %p rather than %d.
3167 81. There was a concern that a line of code in smtp_in.c could overflow a
3168 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3169 initially expressed, the concern was not well-founded, because trailing
3170 spaces are removed early. However, if the trailing spaces were followed by
3171 a NULL, they did not get removed, so the overflow was possible. Two fixes
3174 (a) I re-wrote the offending code in a cleaner fashion.
3175 (b) If an incoming SMTP command contains a NULL character, it is rejected
3178 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3179 runs initgroups(), so that if the Exim user is in any additional groups,
3180 they will be used during message reception.
3186 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob