1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.184 2005/08/01 13:28:30 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.
20 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
22 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
23 command itself ended in error, the underlying message about the transport
24 filter timeout was being overwritten with the pipe command error. Now the
25 underlying error message should be appended to the second error message.
27 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
29 PH/02 Reduce the amount of output that "make" produces by default. Full output
30 can still be requested.
32 PH/03 The warning log line about a condition test deferring for a "warn" verb
33 was being output only once per connection, rather than after each
34 occurrence (because it was using the same function as for successful
35 "warn" verbs). This seems wrong, so I have changed it.
37 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
38 it should not have, which might have caused a crash in the right
39 circumstances, but probably never did.
41 PH/04 Installed a modified version of Tony Finch's patch to make submission
42 mode fix the return path as well as the Sender: header line, and to
43 add a /name= option so that you can make the user's friendly name appear
46 TF/03 Added the control = fakedefer ACL modifier.
48 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
49 Mark Lowes for thorough testing.
51 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
53 TK/03 Merged latest SRS patch from Miles Wilton.
55 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
56 with the definition in sysexits.h (which is #included earlier).
57 Fortunately, Exim does not actually use EX_OK. The code used to try to
58 preserve the sysexits.h value, by assumimg that macro definitions were
59 scanned for macro replacements. I have been disabused of this notion,
60 so now the code just undefines EX_OK before #including unistd.h.
62 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
63 in the smtp transport. When a block could not be written in a single
64 write() function, the timeout was being re-applied to each part-write.
65 This seems wrong - if the receiver was accepting one byte at a time it
66 would take for ever. The timeout is now adjusted when this happens. It
67 doesn't have to be particularly precise.
69 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
70 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
72 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
73 without a domain if the check to <postmaster@domain> fails.
75 SC/01 Eximstats: added -xls and the ability to specify output files
76 (patch written by Frank Heydlauf).
78 SC/02 Eximstats: use FileHandles for outputing results.
80 SC/03 Eximstats: allow any combination of xls, txt, and html output.
82 SC/04 Eximstats: fixed display of large numbers with -nvr option
84 SC/05 Eximstats: fixed merging of reports with empty tables.
86 SC/06 Eximstats: added the -include_original_destination flag
88 SC/07 Eximstats: removed tabs and trailing whitespace.
90 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
92 TK/06 MBOX spool code: Add real "From " MBOX separator line
93 so the .eml file is really in mbox format (even though
94 most programs do not really care). Patch from Alex Miller.
96 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
97 The latter is generated from $received_to and is only set if the
98 message has one envelope recipient. SA can use these headers,
99 obviously out-of-the-box. Patch from Alex Miller.
101 PH/08 The ${def test on a variable was returning false if the variable's
102 value was "0", contrary to what the specification has always said!
103 The result should be true unless the variable is empty.
105 PH/09 The syntax error of a character other than { following "${if
106 def:variable_name" (after optional whitespace) was not being diagnosed.
107 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
108 accidental colon was present, for example, could give incorrect results.
110 PH/10 Tidied the code in a number of places where the st_size field of a stat()
111 result is used (not including appendfile, where other changes are about
114 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
115 This involved changing a lot of size variables from int to off_t. It
116 should work with maildirs and everything.
118 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
119 spamd dying while we are connected to it.
121 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
124 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
125 function for BDB 4.3.
127 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
129 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
130 the sources for PCRE have been re-arranged and more files are now
133 PH/15 The code I had for printing potentially long long variables in PH/11
134 above was not the best (it lost precision). The length of off_t variables
135 is now inspected at build time, and an appropriate printing format (%ld
136 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
137 to be "long long int" or "long int". This is needed for the internal
138 formatting function string_vformat().
140 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
141 the configuration file to be ":syslog", then the script "guesses" where
142 the logs files are, rather than using the compiled in default. In our
143 case the guess is not the same as the compiled default, so the script
144 suddenly stopped working when I started to use syslog. The patch checks
145 to see if log_file_path is "". If so, it attempts to read it from exim
146 with no configuration file to get the compiled in version, before it
147 falls back to the previous guessing code."
149 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
150 implementing BATV in an Exim configuration. See NewStuff for the gory
153 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
154 Makefile that are specific to HP-UX.
156 PH/18 If the "use_postmaster" option was set for a recipient callout together
157 with the "random" option, the postmaster address was used as the MAIL
158 FROM address for the random test, but not for the subsequent recipient
159 test. It is now used for both.
161 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
162 patch removes a few documentation additions to RFC 3028, because the
163 latest draft now contains them. It adds the new en;ascii-case comparator
164 and a new error check for 8bit text in MIME parts. Comparator and
165 require names are now matched exactly. I enabled the subaddress
166 extension, but it is not well tested yet (read: it works for me)."
168 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
169 rework some of the code of TK/09 above to avoid the hardwired use of
170 "%lld" and "long long". Replaced the call to snprintf() with a call to
173 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
174 records point to non-existent hosts", "retry timeout exceeded", and
175 "retry time not reached for any host after a long failure period".
177 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
178 experimental DomainKeys support:
180 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
181 (2) On an error such as an illegally used "control", the wrong name for
182 the control was given.
184 These problems did NOT occur unless DomainKeys support was compiled.
186 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
188 PH/24 Added ${if match_ip condition.
190 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
191 databases so that it will be absolutely obvious if a crash occurs in the
192 DB library. This is a regular occurrence (often caused by mis-matched
195 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
196 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
197 picked up on a user's system that detects such things. There doesn't seem
198 to be a gcc warning option for this - only an attribute that has to be
199 put on the function's prototype. It seems that in Fedora Core 4 they have
200 set this on a number of new functions. No doubt there will be more in due
203 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
204 specify a transport. However, if an address that was verified by such a
205 router was the subject of a callout, Exim crashed because it tried to
206 read the rcpt_include_affixes from the non-existent transport. Now it
207 just assumes that the setting of that option is false. This bug was
208 introduced by 4.51/PH/31.
210 PH/28 Changed -d+all to exclude +memory, because that information is very
211 rarely of interest, but it makes the output a lot bigger. People tend to
212 do -d+all out of habit.
214 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
215 code in os-type was giving problems when libc.so lives in lib64, like on
218 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
219 aren't the modern standard, and it seems that some systems' include files
220 don't always have them. Exim was already checking for some of the newer
221 ones like T_AAAA, and defining it itself. I've added checks for all the
222 record types that Exim uses.
224 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
225 not automatically generating a new one, as it is supposed to. This
226 prevented TLS from working. If the file did exist, but contained invalid
227 data, a new version was generated, as expected. It was only the case of a
228 non-existent file that was broken.
230 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
231 with a change in libdomainkeys > 0.64.
233 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
234 from DNS. If the selector record carries the flag, it now has
235 precedence over the domain-wide flag.
237 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
239 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
240 the use of a port name that isn't defined in /etc/services) occurred, the
241 message was deferred as in a normal delivery, and thus remained on the
242 spool, instead of being failed because of the mua_wrapper setting. This
243 is now fixed, and I tidied up some of the mua_wrapper messages at the
246 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
247 the messages in a hash of arrays rather than using individual hashes.
248 This is a bit cleaner and results in dramatic memory savings, albeit
249 at a slight CPU cost.
251 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
252 as requested by Marc Sherman.
254 SC/10 Eximstats: added histograms for user specified patterns as requested
257 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
259 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
260 fopen() in the content-scanning modules that did not already have it.
266 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
267 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
269 TK/02 Fix ACL "control" statement not being available in MIME ACL.
271 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
273 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
274 to test Sieve filters that use "vacation".
276 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
277 that changes the way the GnuTLS parameters are stored in the cache file.
278 The new format can be generated externally. For backward compatibility,
279 if the data in the cache doesn't make sense, Exim assumes it has read an
280 old-format file, and it generates new data and writes a new file. This
281 means that you can't go back to an older release without removing the
284 PH/03 A redirect router that has both "unseen" and "one_time" set does not
285 work if there are any delivery delays because "one_time" forces the
286 parent to be marked "delivered", so its unseen clone is never tried
287 again. For this reason, Exim now forbids the simultaneous setting of
290 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
291 redirected to themselves ("homonym" addresses). Read the long ChangeLog
292 entry if you want to know the details. The fix, however, neglected to
293 consider the case when local delivery batching is involved. The test for
294 "previously delivered" was not happening when checking to see if an
295 address could be batched with a previous (undelivered) one; under
296 certain circumstances this could lead to multiple deliveries to the same
299 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
300 in its include files, and this causes problems building Exim.
302 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
303 header_syntax) but Exim was just ignoring anything given after a slash.
304 In particular, this caused confusion with an attempt to use "verify =
305 reverse_host_lookup/defer_ok". An error is now given when options are
306 supplied for verify items that do not have them. (Maybe reverse_host_
307 lookup should have a defer_ok option, but that's a different point.)
309 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
310 defined by RFC 821) to 2048, because there were problems with some AUTH
311 commands, and RFC 1869 says the size should be increased for extended
312 SMTP commands that take arguments.
314 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
317 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
318 "unknown" error; now it says that the functionality isn't in the binary.
320 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
321 an address' error message when a string expansion fails (syntax or
322 whatever). Otherwise the password may appear in the log. Following change
323 PH/42 below, there is no longer a chance of it appearing in a bounce
326 PH/11 Installed exipick version 20050225.0 from John Jetmore.
328 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
329 of its addresses was ever tried. (Bugzilla bug #2.)
331 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
332 the result incorrectly in the debug output. (It correctly added a newline
333 to what was transported.)
335 TF/01 Added $received_time.
337 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
338 commented out examples of how to interface to a virus scanner and to
339 SpamAssassin. Also added commented examples of av_scanner and
340 spamd_address settings.
342 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
343 and controls are allowed in which ACLs. There were a couple of minor
344 errors. Some of the entries in the conditions table (which is a table of
345 where they are NOT allowed) were getting very unwieldy; rewrote them as a
346 negation of where the condition IS allowed.
348 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
350 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
351 header file does not have a version number, so I've had to invent a new
352 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
353 API. The code is untested by me (my Linux distribution still has 0.3.2 of
354 radiusclient), but it was contributed by a Radius user.
356 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
357 files or directories, for OpenSSL.
359 PH/19 When an Exim process that is running as root has to create an Exim log
360 file, it does so in a subprocess that runs as exim:exim so as to get the
361 ownership right at creation (otherwise, other Exim processes might see
362 the file with the wrong ownership). There was no test for failure of this
363 fork() call, which would lead to the process getting stuck as it waited
364 for a non-existent subprocess. Forks do occasionally fail when resources
365 run out. I reviewed all the other calls to fork(); they all seem to check
368 PH/20 When checking for unexpected SMTP input at connect time (before writing
369 the banner), Exim was not dealing correctly with a non-positive return
370 from the read() function. If the client had disconnected by this time,
371 the result was a log entry for a synchronization error with an empty
372 string after "input=" when read() returned zero. If read() returned -1
373 (an event I could not check), uninitialized data bytes were printed.
374 There were reports of junk text (parts of files, etc) appearing after
377 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
379 PH/22 Added support for macro redefinition, and (re)definition in between
380 driver and ACL definitions.
382 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
383 forgetting to use the resulting value; it was using the unexpanded value.
385 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
386 hadn't been configured. The fix is from Juergen Kreileder, who
387 understands it better than I do:
389 "Here's what I see happening with three configured cyrus_sasl
390 authenticators configured (plain, login, cram-md5):
392 On startup auth_cyrus_sasl_init() gets called for each of these.
393 This means three calls to sasl_listmech() without a specified mech_list.
394 => SASL tests which mechs of all available mechs actually work
395 => three warnings about OTP not working
396 => the returned list contains: plain, login, cram-md5, digest-md5, ...
398 With the patch, sasl_listmech() also gets called three times. But now
399 SASL's mech_list option is set to the server_mech specified in the the
400 authenticator. Or in other words, the answer from sasl_listmech()
401 gets limited to just the mech you're testing for (which is different
403 => the return list contains just 'plain' or 'login', 'cram-md5' or
404 nothing depending on the value of ob->server_mech.
406 I've just tested the patch: Authentication still works fine,
407 unavailable mechs specified in the exim configuration are still
408 caught, and the auth.log warnings about OTP are gone."
410 PH/25 When debugging is enabled, the contents of the command line are added
411 to the debugging output, even when log_selector=+arguments is not
414 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
415 answer is "GNU", and only if the return is "GNU/something" is the answer
418 PH/27 $acl_verify_message is now set immediately after the failure of a
419 verification in an ACL, and so is available in subsequent modifiers. In
420 particular, the message can be preserved by coding like this:
422 warn !verify = sender
423 set acl_m0 = $acl_verify_message
425 Previously, $acl_verify_message was set only while expanding "message"
426 and "log_message" when a very denied access.
428 PH/28 Modified OS/os.c-Linux with
430 -#ifndef OS_LOAD_AVERAGE
431 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
433 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
434 nomenclature these days.)
436 PH/29 Installed patch from the Sieve maintainer that adds the options
437 sieve_useraddress and sieve_subaddress to the redirect router.
439 PH/30 In these circumstances:
440 . Two addresses routed to the same list of hosts;
441 . First host does not offer TLS;
442 . First host accepts first address;
443 . First host gives temporary error to second address;
444 . Second host offers TLS and a TLS session is established;
445 . Second host accepts second address.
446 Exim incorrectly logged both deliveries with the TLS parameters (cipher
447 and peerdn, if requested) that were in fact used only for the second
450 PH/31 When doing a callout as part of verifying an address, Exim was not paying
451 attention to any local part prefix or suffix that was matched by the
452 router that accepted the address. It now behaves in the same way as it
453 does for delivery: the affixes are removed from the local part unless
454 rcpt_include_affixes is set on the transport.
456 PH/32 Add the sender address, as F=<...>, to the log line when logging a
457 timeout during the DATA phase of an incoming message.
459 PH/33 Sieve envelope tests were broken for match types other than :is. I have
460 applied a patch sanctioned by the Sieve maintainer.
462 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
463 the uid or gid is negative. A case of a negative gid caused this to be
464 noticed. The fix allows for either to be negative.
466 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
467 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
470 PH/36 The change PH/12 above was broken. Fixed it.
472 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
473 the grounds that routing the same address twice would always produce the
474 same answer. This might have been true once, but it is certainly no
475 longer true now. Routing a child address may depend on the previous
476 routing that produced that child. Some complicated redirection strategies
477 went wrong when messages had multiple recipients, and made Exim's
478 behaviour dependent on the order in which the addresses were given.
480 I have moved the duplicate checking until after the routing is complete.
481 Exim scans the addresses that are assigned to local and remote
482 transports, and removes any duplicates. This means that more work will be
483 done, as duplicates will always all be routed, but duplicates are
484 presumably rare, so I don't expect this is of any significance.
486 For deliveries to pipes, files, and autoreplies, the duplicate checking
487 still happens during the routing process, since they are not going to be
490 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
491 It corrects a timeout issue with spamd. This is Ian's comment: "The
492 background is that sometimes spamd either never reads data from a
493 connection it has accepted, or it never writes response data. The exiscan
494 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
495 blindly assumes that writes won't block so it may never time out."
497 PH/39 Allow G after quota size as well as K and M.
499 PH/40 The value set for $authenticated_id in an authenticator may not contain
500 binary zeroes or newlines because the value is written to log lines and
501 to spool files. There was no check on this. Now the value is run through
502 the string_printing() function so that such characters are converted to
503 printable escape sequences.
505 PH/41 $message_linecount is a new variable that contains the total number of
506 lines in the message. Compare $body_linecount, which is the count for the
509 PH/42 Exim no longer gives details of delivery errors for specific addresses in
510 bounce and delay warning messages, except in certain special cases, which
513 (a) An SMTP error message from a remote host;
514 (b) A message specified in a :fail: redirection;
515 (c) A message specified in a "fail" command in a system filter;
516 (d) A message specified in a FAIL return from the queryprogram router;
517 (e) A message specified by the cannot_route_message router option.
519 In these cases only, Exim does include the error details in bounce and
520 warning messages. There are also a few cases where bland messages such
521 as "unrouteable address" or "local delivery error" are given.
523 PH/43 $value is now also set for the "else" part of a ${run expansion.
525 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
526 being worked on, but at least Exim now implements the latest version to
529 PH/45 In a pipe transport, although a timeout while waiting for the pipe
530 process to complete was treated as a delivery failure, a timeout while
531 writing the message to the pipe was logged, but erroneously treated as a
532 successful delivery. Such timeouts include transport filter timeouts. For
533 consistency with the overall process timeout, these timeouts are now
534 treated as errors, giving rise to delivery failures by default. However,
535 there is now a new Boolean option for the pipe transport called
536 timeout_defer, which, if set TRUE, converts the failures into defers for
537 both kinds of timeout. A transport filter timeout is now identified in
540 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
541 systems where "make" and "gmake" are different, calling "gmake" at top
542 level broke things. I've arranged for the value of $(MAKE) to be passed
543 from the Makefile to this script so that it can call the same version of
547 A note about Exim versions 4.44 and 4.50
548 ----------------------------------------
550 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
551 changes of various kinds. As a consequence, a big documentation update was
552 needed. This delayed the release for rather longer than seemed good, especially
553 in the light of a couple of (minor) security issues. Therefore, the changes
554 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
555 release. So 4.44 and 4.50 are in effect two different branches that both start
558 I have left the 4.50 change log unchanged; it contains all the changes since
559 4.43. The change log for 4.44 is below; many of its items are identical to
560 those for 4.50. This seems to be the most sensible way to preserve the
561 historical information.
567 1. Minor wording change to the doc/README.SIEVE file.
569 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
570 computation of the current number of files was incorrect.
572 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
573 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
576 4. Give more explanation in the error message when the command for a transport
577 filter fails to execute.
579 5. There are several places where Exim runs a non-Exim command in a
580 subprocess. The SIGUSR1 signal should be disabled for these processes. This
581 was being done only for the command run by the queryprogram router. It is
582 now done for all such subprocesses. The other cases are: ${run, transport
583 filters, and the commands run by the lmtp and pipe transports.
585 6. Added CONFIGURE_GROUP build-time option.
587 7. Some older OS have a limit of 256 on the maximum number of file
588 descriptors. Exim was using setrlimit() to set 1000 as a large value
589 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
590 systems. I've change it so that if it can't get 1000, it tries for 256.
592 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
593 was an oversight, and furthermore, ever since the addition of extra
594 controls (e.g. 4.43/32), the checks on when to allow different forms of
595 "control" were broken. There should now be diagnostics for all cases when a
596 control that does not make sense is encountered.
598 9. Added the /retain_sender option to "control=submission".
600 10. $recipients is now available in the predata ACL (oversight).
602 11. Tidy the search cache before the fork to do a delivery from a message
603 received from the command line. Otherwise the child will trigger a lookup
604 failure and thereby defer the delivery if it tries to use (for example) a
605 cached ldap connection that the parent has called unbind on.
607 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
608 of $address_data from the recipient verification was clobbered by the
611 13. The value of address_data from a sender verification is now available in
612 $sender_address_data in subsequent conditions in the ACL statement.
614 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
616 15. Added a new option "connect=<time>" to callout options, to set a different
619 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
620 was its contents. (It was OK if the option was not defined at all.)
622 17. A "Completed" log line is now written for messages that are removed from
623 the spool by the -Mrm option.
625 18. New variables $sender_verify_failure and $recipient_verify_failure contain
626 information about exactly what failed.
628 19. Added -dd to debug only the daemon process.
630 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
631 handles timeouts, both on the server side and network timeouts. Renamed the
632 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
634 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
635 It is now set to "smtps".
637 22. $host_address is now set to the target address during the checking of
640 23. When checking ignore_target_hosts for an ipliteral router, no host name was
641 being passed; this would have caused $sender_host_name to have been used if
642 matching the list had actually called for a host name (not very likely,
643 since this list is usually IP addresses). A host name is now passed as
646 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
647 code that specifies a non-restarting handler (typically sigaction() in
648 modern systems) in an attempt to fix a rare and obscure crash bug.
650 25. Narrowed the window for a race in the daemon that could cause it to ignore
651 SIGCHLD signals. This is not a major problem, because they are used only to
652 wake it up if nothing else does.
654 26. A malformed maildirsize file could cause Exim to calculate negative values
655 for the mailbox size or file count. Odd effects could occur as a result.
656 The maildirsize information is now recalculated if the size or filecount
659 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
660 support for a long time. Removed HAVE_SYS_VFS_H.
662 28. Installed the latest version of exipick from John Jetmore.
664 29. In an address list, if the pattern was not a regular expression, an empty
665 subject address (from a bounce message) matched only if the pattern was an
666 empty string. Non-empty patterns were not even tested. This was the wrong
667 because it is perfectly reasonable to use an empty address as part of a
668 database query. An empty address is now tested by patterns that are
669 lookups. However, all the other forms of pattern expect the subject to
670 contain a local part and a domain, and therefore, for them, an empty
671 address still always fails if the pattern is not itself empty.
673 30. Exim went into a mad DNS loop when attempting to do a callout where the
674 host was specified on an smtp transport, and looking it up yielded more
677 31. Re-factored the code for checking spool and log partition space into a
678 function that finds that data and another that does the check. The former
679 is then used to implement four new variables: $spool_space, $log_space,
680 $spool_inodes, and $log_inodes.
682 32. The RFC2047 encoding function was originally intended for short strings
683 such as real names; it was not keeping to the 75-character limit for
684 encoded words that the RFC imposes. It now respects the limit, and
685 generates multiple encoded words if necessary. To be on the safe side, I
686 have increased the buffer size for the ${rfc2047: expansion operator from
689 33. It is now permitted to omit both strings after an "if" condition; if the
690 condition is true, the result is "true". As before, when the second string
691 is omitted, a false condition yields an empty string. This makes it less
692 cumbersome to write custom ACL and router conditions.
694 34. Failure to deliver a bounce message always caused it to be frozen, even if
695 there was an errors_to setting on the router. The errors_to setting is now
698 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
699 canonical form (fully expanded) before being placed in
700 $sender_host_address.
702 36. The table in the code that translates DNS record types into text (T_A to
703 "A" for instance) was missing entries for NS and CNAME. It is just possible
704 that this could have caused confusion if both these types were looked up
705 for the same domain, because the text type is used as part of Exim's
706 per-process caching. But the chance of anyone hitting this buglet seems
709 37. The dnsdb lookup has been extended in a number of ways.
711 (1) There is a new type, "zns", which walks up the domain tree until it
712 finds some nameserver records. It should be used with care.
714 (2) There is a new type, "mxh", which is like "mx" except that it returns
715 just the host names, not the priorities.
717 (3) It is now possible to give a list of domains (or IP addresses) to be
718 looked up. The behaviour when one of the lookups defers can be
719 controlled by a keyword.
721 (4) It is now possible to specify the separator character for use when
722 multiple records are returned.
724 38. The dnslists ACL condition has been extended: it is now possible to supply
725 a list of IP addresses and/or domains to be looked up in a particular DNS
728 39. Added log_selector=+queue_time_overall.
730 40. When running the queue in the test harness, wait just a tad after forking a
731 delivery process, to get repeatability of debugging output.
733 41. Include certificate and key file names in error message when GnuTLS fails
734 to set them up, because the GnuTLS error message doesn't include the name
735 of the failing file when there is a problem reading it.
737 42. Allow both -bf and -bF in the same test run.
739 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
741 44. The "Exiscan patch" is now merged into the mainline Exim source.
743 45. Sometimes the final signoff response after QUIT could fail to get
744 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
745 < 0 before doing a fflush(). This bug looks as though it goes back to the
746 introduction of TLS in release 3.20, but "sometimes" must have been rare
747 because the tests only now provoked it.
749 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
750 (this can affect the format of dates).
752 47. exim_tidydb, when checking for the continued existence of a message for
753 which it has found a message-specific retry record, was not finding
754 messages that were in split spool directories. Consequently, it was
755 deleting retry records that should have stayed in existence.
757 48. Steve fixed some bugs in eximstats.
759 49. The SPA authentication driver was not abandoning authentication and moving
760 on to the next authenticator when an expansion was forced to fail,
761 contradicting the general specification for all authenticators. Instead it
762 was generating a temporary error. It now behaves as specified.
764 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
765 (the order specifies the preference for clients). The order is now AES256,
766 AES128, 3DES, ARCFOUR128.
768 51. Small patch to Sieve code - explicitly set From: when generating an
771 52. Exim crashed if a remote delivery caused a very long error message to be
772 recorded - for instance if somebody sent an entire SpamAssassin report back
773 as a large number of 550 error lines. This bug was coincidentally fixed by
774 increasing the size of one of Exim's internal buffers (big_buffer) that
775 happened as part of the Exiscan merge. However, to be on the safe side, I
776 have made the code more robust (and fixed the comments that describe what
779 53. Now that there can be additional text after "Completed" in log lines (if
780 the queue_time_overall log selector is set), a one-byte patch to exigrep
781 was needed to allow it to recognize "Completed" as not the last thing in
784 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
785 patch that reportedly fixes this has been added. I am not expert enough to
786 create a test for it. This is what the patch creator wrote:
788 "I found a little strange behaviour of ldap code when working with
789 Windows 2003 AD Domain, where users was placed in more than one
790 Organization Units. When I tried to give exim partial DN, the exit code
791 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
792 But simultaneously result of request was absolutely normal ldap result,
793 so I produce this patch..."
795 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
796 REFERENCE, so I have modified the code to exclude the patch when that macro
799 55. Some experimental protocols are using DNS PTR records for new purposes. The
800 keys for these records are domain names, not reversed IP addresses. The
801 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
802 leaves it alone. Component reversal etc. now happens only for IP addresses.
805 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
807 57. Double the size of the debug message buffer (to 2048) so that more of very
808 long debug lines gets shown.
810 58. The exicyclog utility now does better if the number of log files to keep
811 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
813 59. Two changes related to the smtp_active_hostname option:
815 (1) $smtp_active_hostname is now available as a variable.
816 (2) The default for smtp_banner uses $smtp_active_hostname instead
817 of $primary_hostname.
819 60. The host_aton() function is supposed to be passed a string that is known
820 to be a valid IP address. However, in the case of IPv6 addresses, it was
821 not checking this. This is a hostage to fortune. Exim now panics and dies
822 if the condition is not met. A case was found where this could be provoked
823 from a dnsdb PTR lookup with an IPv6 address that had more than 8
824 components; fortuitously, this particular loophole had already been fixed
825 by change 4.50/55 above.
827 If there are any other similar loopholes, the new check in host_aton()
828 itself should stop them being exploited. The report I received stated that
829 data on the command line could provoke the exploit when Exim was running as
830 exim, but did not say which command line option was involved. All I could
831 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
835 61. There was a buffer overflow vulnerability in the SPA authentication code
836 (which came originally from the Samba project). I have added a test to the
837 spa_base64_to_bits() function which I hope fixes it.
840 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
841 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
843 63. The daemon start-up calls getloadavg() while still root for those OS that
844 need the first call to be done as root, but it missed one case: when
845 deliver_queue_load_max is set with deliver_drop_privilege. This is
846 necessary for the benefit of the queue runner, because there is no re-exec
847 when deliver_drop_privilege is set.
849 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
852 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
853 was in use, was not putting the data itself into the right store pool;
854 consequently, it could be overwritten for a subsequent message in the same
855 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
858 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
860 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
861 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
862 the function were treating the return as a boolean value, which happened to
863 work because 0=false and not-0=true, but is not correct code.
865 68. The host_aton() function was not handling scoped IPv6 addresses (those
866 with, for example, "%eth0" on the end) correctly.
868 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
869 negated items (that is, ~something) in unsigned ints. Some compilers
870 apparently mutter when there is no cast.
872 70. If an address verification called from an ACL failed, and did not produce a
873 user-specific message (i.e. there was only a "system" message), nothing was
874 put in $acl_verify_message. In this situation, it now puts the system
877 71. Change 4.23/11 added synchronization checking at the start of an SMTP
878 session; change 4.31/43 added the unwanted input to the log line - except
879 that it did not do this in the start of session case. It now does.
881 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
882 This is wrong and can cause the other end to generate a synchronization
883 error if it is another Exim or anything else that does the synchronization
884 check. A QUIT command is no longer sent after a timeout.
886 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
889 74. The defer_ok option of callout verification was not working if it was used
890 when verifying addresses in header lines, that is, for this case:
892 verify = header_sender/callout=defer_ok
894 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
895 those file descriptors could be used for SMTP connections. If anything
896 wrote to stderr (the example that came up was "warn" in embedded Perl), it
897 could be sent to the SMTP client, causing chaos. The daemon now opens
898 stdin, stdout, and stderr to /dev/null when it puts itself into the
901 76. Arrange for output from Perl's "warn" command to be written to Exim's main
902 log by default. The user can override this with suitable Perl magic.
904 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
905 the log message when discard triggers, was not working for the DATA ACL or
906 for the non-SMTP ACL.
908 78. Error message wording change in sieve.c.
910 79. If smtp_accept_max_per_host was set, the number of connections could be
911 restricted to fewer than expected, because the daemon was trying to set up
912 a new connection before checking whether the processes handling previous
913 connections had finished. The check for completed processes is now done
914 earlier. On busy systems, this bug wouldn't be noticed because something
915 else would have woken the daemon, and it would have reaped the completed
918 80. If a message was submitted locally by a user whose login name contained one
919 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
920 It caused a spool format error. I have fixed the spool reading code. A
921 related problem was that the "from" clause in the Received: line became
922 illegal because of the space(s). It is now covered by ${quote_local_part.
924 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
927 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
929 83. Patch from Sieve maintainer to fix unterminated string problem in
932 84. Some minor changes to the Linux configuration files to help with other
933 OS variants using glibc.
935 85. One more patch for Sieve to update vacation handling to latest spec.
938 ----------------------------------------------------
939 See the note above about the 4.44 and 4.50 releases.
940 ----------------------------------------------------
946 1. Change 4.43/35 introduced a bug that caused file counts to be
947 incorrectly computed when quota_filecount was set in an appendfile
950 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
951 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
954 3. Give more explanation in the error message when the command for a transport
955 filter fails to execute.
957 4. There are several places where Exim runs a non-Exim command in a
958 subprocess. The SIGUSR1 signal should be disabled for these processes. This
959 was being done only for the command run by the queryprogram router. It is
960 now done for all such subprocesses. The other cases are: ${run, transport
961 filters, and the commands run by the lmtp and pipe transports.
963 5. Some older OS have a limit of 256 on the maximum number of file
964 descriptors. Exim was using setrlimit() to set 1000 as a large value
965 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
966 systems. I've change it so that if it can't get 1000, it tries for 256.
968 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
969 was an oversight, and furthermore, ever since the addition of extra
970 controls (e.g. 4.43/32), the checks on when to allow different forms of
971 "control" were broken. There should now be diagnostics for all cases when a
972 control that does not make sense is encountered.
974 7. $recipients is now available in the predata ACL (oversight).
976 8. Tidy the search cache before the fork to do a delivery from a message
977 received from the command line. Otherwise the child will trigger a lookup
978 failure and thereby defer the delivery if it tries to use (for example) a
979 cached ldap connection that the parent has called unbind on.
981 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
982 of $address_data from the recipient verification was clobbered by the
985 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
986 was its contents. (It was OK if the option was not defined at all.)
988 11. A "Completed" log line is now written for messages that are removed from
989 the spool by the -Mrm option.
991 12. $host_address is now set to the target address during the checking of
994 13. When checking ignore_target_hosts for an ipliteral router, no host name was
995 being passed; this would have caused $sender_host_name to have been used if
996 matching the list had actually called for a host name (not very likely,
997 since this list is usually IP addresses). A host name is now passed as
1000 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1001 code that specifies a non-restarting handler (typically sigaction() in
1002 modern systems) in an attempt to fix a rare and obscure crash bug.
1004 15. Narrowed the window for a race in the daemon that could cause it to ignore
1005 SIGCHLD signals. This is not a major problem, because they are used only to
1006 wake it up if nothing else does.
1008 16. A malformed maildirsize file could cause Exim to calculate negative values
1009 for the mailbox size or file count. Odd effects could occur as a result.
1010 The maildirsize information is now recalculated if the size or filecount
1013 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1014 support for a long time. Removed HAVE_SYS_VFS_H.
1016 18. Updated exipick to current release from John Jetmore.
1018 19. Allow an empty sender to be matched against a lookup in an address list.
1019 Previously the only cases considered were a regular expression, or an
1022 20. Exim went into a mad DNS lookup loop when doing a callout where the
1023 host was specified on the transport, if the DNS lookup yielded more than
1026 21. The RFC2047 encoding function was originally intended for short strings
1027 such as real names; it was not keeping to the 75-character limit for
1028 encoded words that the RFC imposes. It now respects the limit, and
1029 generates multiple encoded words if necessary. To be on the safe side, I
1030 have increased the buffer size for the ${rfc2047: expansion operator from
1033 22. Failure to deliver a bounce message always caused it to be frozen, even if
1034 there was an errors_to setting on the router. The errors_to setting is now
1037 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1038 canonical form (fully expanded) before being placed in
1039 $sender_host_address.
1041 24. Updated eximstats to version 1.33
1043 25. Include certificate and key file names in error message when GnuTLS fails
1044 to set them up, because the GnuTLS error message doesn't include the name
1045 of the failing file when there is a problem reading it.
1047 26. Expand error message when OpenSSL has problems setting up cert/key files.
1050 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1051 (this can affect the format of dates).
1053 28. exim_tidydb, when checking for the continued existence of a message for
1054 which it has found a message-specific retry record, was not finding
1055 messages that were in split spool directories. Consequently, it was
1056 deleting retry records that should have stayed in existence.
1058 29. eximstats updated to version 1.35
1059 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1060 1.35 - bugfix such that pie charts by volume are generated correctly
1062 30. The SPA authentication driver was not abandoning authentication and moving
1063 on to the next authenticator when an expansion was forced to fail,
1064 contradicting the general specification for all authenticators. Instead it
1065 was generating a temporary error. It now behaves as specified.
1067 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1068 (the order specifies the preference for clients). The order is now AES256,
1069 AES128, 3DES, ARCFOUR128.
1071 31. Small patch to Sieve code - explicitly set From: when generating an
1074 32. Exim crashed if a remote delivery caused a very long error message to be
1075 recorded - for instance if somebody sent an entire SpamAssassin report back
1076 as a large number of 550 error lines. This bug was coincidentally fixed by
1077 increasing the size of one of Exim's internal buffers (big_buffer) that
1078 happened as part of the Exiscan merge. However, to be on the safe side, I
1079 have made the code more robust (and fixed the comments that describe what
1082 33. Some experimental protocols are using DNS PTR records for new purposes. The
1083 keys for these records are domain names, not reversed IP addresses. The
1084 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1085 leaves it alone. Component reversal etc. now happens only for IP addresses.
1088 34. The host_aton() function is supposed to be passed a string that is known
1089 to be a valid IP address. However, in the case of IPv6 addresses, it was
1090 not checking this. This is a hostage to fortune. Exim now panics and dies
1091 if the condition is not met. A case was found where this could be provoked
1092 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1093 components; fortuitously, this particular loophole had already been fixed
1094 by change 4.50/55 or 4.44/33 above.
1096 If there are any other similar loopholes, the new check in host_aton()
1097 itself should stop them being exploited. The report I received stated that
1098 data on the command line could provoke the exploit when Exim was running as
1099 exim, but did not say which command line option was involved. All I could
1100 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1101 running as the user.
1104 35. There was a buffer overflow vulnerability in the SPA authentication code
1105 (which came originally from the Samba project). I have added a test to the
1106 spa_base64_to_bits() function which I hope fixes it.
1109 36. The daemon start-up calls getloadavg() while still root for those OS that
1110 need the first call to be done as root, but it missed one case: when
1111 deliver_queue_load_max is set with deliver_drop_privilege. This is
1112 necessary for the benefit of the queue runner, because there is no re-exec
1113 when deliver_drop_privilege is set.
1115 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1116 was in use, was not putting the data itself into the right store pool;
1117 consequently, it could be overwritten for a subsequent message in the same
1118 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1121 38. Sometimes the final signoff response after QUIT could fail to get
1122 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1123 < 0 before doing a fflush(). This bug looks as though it goes back to the
1124 introduction of TLS in release 3.20, but "sometimes" must have been rare
1125 because the tests only now provoked it.
1131 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1132 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1133 Now it returns an integer. A number of calls were still expecting a T/F
1134 return. Fortuitously, in all cases, the tests worked in OK situations,
1135 which is the norm. However, things would have gone wrong on any write
1136 failures on the smtp file descriptor. This function is used when sending
1137 messages over SMTP and also when doing verify callouts.
1139 2. When Exim is called to do synchronous delivery of a locally submitted
1140 message (the -odf or -odi options), it no longer closes stderr before doing
1143 3. Implemented the mua_wrapper option.
1145 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1147 5. Implemented the functions header_remove(), header_testname(),
1148 header_add_at_position(), and receive_remove_recipient(), and exported them
1151 6. If an ACL "warn" statement specified the addition of headers, Exim already
1152 inserted X-ACL-Warn: at the start if there was no header name. However, it
1153 was not making this test for the second and subsequent header lines if
1154 there were newlines in the string. This meant that an invalid header could
1155 be inserted if Exim was badly configured.
1157 7. Allow an ACL "warn" statement to add header lines at the start or after all
1158 the Received: headers, as well as at the end.
1160 8. Added the rcpt_4xx retry error code.
1162 9. Added postmaster_mailfrom=xxx to callout verification option.
1164 10. Added mailfrom=xxxx to the callout verification option, for verify=
1167 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1168 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1170 12. Inserted some casts to stop certain compilers warning when using pointer
1171 differences as field lengths or precisions in printf-type calls (mostly
1172 affecting debugging statements).
1174 13. Added optional readline() support for -be (dynamically loaded).
1176 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1177 same clock tick as a message's arrival, so that its received time was the
1178 same as the "first fail" time on the retry record, and that message
1179 remained on the queue past the ultimate address timeout, every queue runner
1180 would try a delivery (because it was past the ultimate address timeout) but
1181 after another failure, the ultimate address timeout, which should have then
1182 bounced the address, did not kick in. This was a "< instead of <=" error;
1183 in most cases the first failure would have been in the next clock tick
1184 after the received time, and all would be well.
1186 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1187 being recognized when the domain list was tested by the match_domain
1188 condition in an expansion string.
1190 16. Added the ${str2b64: operator.
1192 17. Exim was always calling setrlimit() to set a large limit for the number of
1193 processes, without checking whether the existing limit was already
1194 adequate. (It did check for the limit on file descriptors.) Furthermore,
1195 errors from getrlimit() and setrlimit() were being ignored. Now they are
1196 logged to the main and panic logs, but Exim does carry on, to try to do its
1197 job under whatever limits there are.
1199 18. Imported PCRE 5.0.
1201 19. Trivial typo in log message " temporarily refused connection" (the leading
1204 20. If the log selector return_path_on_delivery was set and an address was
1205 redirected to /dev/null, the delivery process crashed because it assumed
1206 that a return path would always be set for a "successful" delivery. In this
1207 case, the whole delivery is bypassed as an optimization, and therefore no
1210 21. Internal re-arrangement: the function for sending a challenge and reading
1211 a response while authentication was assuming a zero-terminated challenge
1212 string. It's now changed to take a pointer and a length, to allow for
1213 binary data in such strings.
1215 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1217 23. Exim was not respecting finduser_retries when seeking the login of the
1218 uid under which it was called; it was always trying 10 times. (The default
1219 setting of finduser_retries is zero.) Also, it was sleeping after the final
1220 failure, which is pointless.
1222 24. Implemented tls_on_connect_ports.
1224 25. Implemented acl_smtp_predata.
1226 26. If the domain in control=submission is set empty, Exim assumes that the
1227 authenticated id is a complete email address when it generates From: or
1228 Sender: header lines.
1230 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1231 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1232 chown and chgrp in /bin and hostname in /usr/bin.
1234 28. Exim was keeping the "process log" file open after each use, just as it
1235 does for the main log. This opens the possibility of it remaining open for
1236 long periods when the USR1 signal hits a daemon. Occasional processlog
1237 errors were reported, that could have been caused by this. Anyway, it seems
1238 much more sensible not to leave this file open at all, so that is what now
1241 29. The long-running daemon process does not normally write to the log once it
1242 has entered its main loop, and it closes the log before doing so. This is
1243 so that log files can straightforwardly be renamed and moved. However,
1244 there are a couple of unusual error situations where the daemon does write
1245 log entries, and I had neglected to close the log afterwards.
1247 30. The text of an SMTP error response that was received during a remote
1248 delivery was being truncated at 512 bytes. This is too short for some of
1249 the long messages that one sometimes sees. I've increased the limit to
1252 31. It is now possible to make retry rules that apply only when a message has a
1253 specific sender, in particular, an empty sender.
1255 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1256 it possible to be selective about when SMTP synchronization is enforced.
1258 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1260 32. Implemented hosts_connection_nolog.
1262 33. Added an ACL for QUIT.
1264 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1267 35. Added mailbox_size and mailbox_filecount to appendfile.
1269 36. Added control = no_multiline_responses to ACLs.
1271 37. There was a bug in the logic of the code that waits for the clock to tick
1272 in the case where the clock went backwards by a substantial amount such
1273 that the microsecond fraction of "now" was more than the microsecond
1274 fraction of "then" (but the whole seconds number was less).
1276 38. Added support for the libradius Radius client library this is found on
1277 FreeBSD (previously only the radiusclient library was supported).
1283 1. When certain lookups returned multiple values in the form name=value, the
1284 quoting of the values was not always being done properly. Specifically:
1285 (a) If the value started with a double quote, but contained no whitespace,
1287 (b) If the value contained whitespace other than a space character (i.e.
1288 tabs or newlines or carriage returns) it was not quoted.
1289 This fix has been applied to the mysql and pgsql lookups by writing a
1290 separate quoting function and calling it from the lookup code. The fix
1291 should probably also be applied to nisplus, ibase and oracle lookups, but
1292 since I cannot test any of those, I have not disturbed their existing code.
1294 2. A hit in the callout cache for a specific address caused a log line with no
1295 reason for rejecting RCPT. Now it says "Previous (cached) callout
1296 verification failure".
1298 3. There was an off-by-one bug in the queryprogram router. An over-long
1299 return line was truncated at 256 instead of 255 characters, thereby
1300 overflowing its buffer with the terminating zero. As well as fixing this, I
1301 have increased the buffer size to 1024 (and made a note to document this).
1303 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1304 when Exim is waiting for an SMTP response from a remote server, Exim
1305 restarts its select() call on the socket, thereby resetting its timeout.
1306 This is not a problem when such interrupts are rare. Somebody set up a cron
1307 job to run exiwhat every 2 minutes, which is less than the normal select()
1308 timeout (5 or 10 minutes). This meant that the select() timeout never
1309 kicked in because it was always reset. I have fixed this by comparing the
1310 time when an interrupt arrives with the time at the start of the first call
1311 to select(). If more time than the timeout has elapsed, the interrupt is
1312 treated as a timeout.
1314 5. Some internal re-factoring in preparation for the addition of Sieve
1315 extensions (by MH). In particular, the "personal" test is moved to a
1316 separate function, and given an option for scanning Cc: and Bcc: (which is
1317 not set for Exim filters).
1319 6. When Exim created an email address using the login of the caller as the
1320 local part (e.g. when creating a From: or Sender: header line), it was not
1321 quoting the local part when it contained special characters such as @.
1323 7. Installed new OpenBSD configuration files.
1325 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1326 try to make them clearer.
1328 9. Callout options, other than the timeout value, were being ignored when
1329 verifying sender addresses in header lines. For example, when using
1331 verify = header_sender/callout=no_cache
1333 the cache was (incorrectly) being used.
1335 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1336 only the Cygwin environment.
1338 11. When return_path_on_delivery was set as a log selector, if different remote
1339 addresses in the same message used different return paths and parallel
1340 remote delivery occurred, the wrong values would sometimes be logged.
1341 (Whenever a remote delivery process finished, the return path value from
1342 the most recently started remote delivery process was logged.)
1344 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1345 lines when AUTH and/or TLS are in use. This is the "received protocol"
1346 field. Exim used to use "asmtp" for authenticated SMTP, without any
1347 indication (in the protocol name) for TLS use. Now it follows the RFC and
1348 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1349 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1350 names appear in log lines as well as in Received: header lines.
1352 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1353 extensions, and comparison tests, and to fix some bugs.
1355 14. Changes to the "personal" filter test:
1357 (1) The test was buggy in that it was just doing the equivalent of
1358 "contains" tests on header lines. For example, if a user's address was
1359 anne@some.where, the "personal" test would incorrectly be true for
1361 To: susanne@some.where
1363 This test is now done by extracting each address from the header in turn,
1364 and checking the entire address. Other tests that are part of "personal"
1365 are now done using regular expressions (for example, to check local parts
1366 of addresses in From: header lines).
1368 (2) The list of non-personal local parts in From: addresses has been
1369 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1370 taken from the Sieve specification recommendations.
1372 (3) If the message contains any header line starting with "List-" it is
1373 treated as non-personal.
1375 (4) The test for "circular" in the Subject: header line has been removed
1376 because it now seems ill-conceived.
1378 15. Minor typos in src/EDITME comments corrected.
1380 16. Installed latest exipick from John Jetmore.
1382 17. If headers_add on a router specified a text string that was too long for
1383 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1384 of string_sprintf() is now avoided.
1386 18. $message_body_size was not set (it was always zero) when running the DATA
1387 ACL and the local_scan() function.
1389 19. For the "mail" command in an Exim filter, no default was being set for
1390 the once_repeat time, causing a random time value to be used if "once" was
1391 specified. (If the value happened to be <= 0, no repeat happened.) The
1392 default is now 0s, meaning "never repeat". The "vacation" command was OK
1393 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1394 (I found it when inspecting the code).
1396 20. There is now an overall timeout for performing a callout verification. It
1397 defaults to 4 times the callout timeout, which applies to individual SMTP
1398 commands during the callout. The overall timeout applies when there is more
1399 than one host that can be tried. The timeout is checked before trying the
1400 next host. This prevents very long delays if there are a large number of
1401 hosts and all are timing out (e.g. when the network connections are timing
1402 out). The value of the overall timeout can be changed by specifying an
1403 additional sub-option for "callout", called "maxwait". For example:
1405 verify = sender/callout=5s,maxwait=20s
1407 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1408 to the end before writing, but this should make it even safer).
1410 22. Exim was forgetting that it had advertised PIPELINING for the second and
1411 subsequent messages on an SMTP connection. It was also not resetting its
1412 memory on STARTTLS and an internal HELO.
1414 23. When Exim logs an SMTP synchronization error within a session, it now
1415 records whether PIPELINING has been advertised or not.
1417 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1418 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1419 rather than long int.
1421 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1423 26. Added the never_mail option to autoreply.
1429 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1430 crash if the getsockname() call failed; this can happen if a connection is
1431 closed very soon after it is established. The problem was simply in the
1432 order in which certain operations were done, causing Exim to try to write
1433 to the SMTP stream before it had set up the file descriptor. The bug has
1434 been fixed by making things happen in the correct order.
1440 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1441 before the connection was closed, thus losing the rejection response.
1443 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1444 some early Solaris releases, but causes trouble in current releases where
1445 socklen_t is defined.
1447 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1450 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1453 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1454 that releasing the top of it at the end releases what was used for sub-
1455 expansions (unless the block got too big). However, discard this block if
1456 the first thing is a variable or header, so that we can use its block when
1457 it is dynamic (useful for very large $message_headers, for example).
1459 6. Lookups now cache *every* query, not just the most recent. A new, separate
1460 store pool is used for this. It can be recovered when all lookup caches are
1461 flushed. Lookups now release memory at the end of their result strings.
1462 This has involved some general refactoring of the lookup sources.
1464 7. Some code has been added to the store_xxx() functions to reduce the amount
1465 of flapping under certain conditions.
1467 8. log_incoming_interface used to affect only the <= reception log lines. Now
1468 it causes the local interface and port to be added to several more SMTP log
1469 lines, for example "SMTP connection from", and rejection lines.
1471 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1473 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1475 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1476 could be overwritten at the end of the current message (or the start of a
1477 new message if it was set in a HELO ACL). The value is now preserved for
1478 the duration of the SMTP connection.
1480 12. If a transport had a headers_rewrite setting, and a matching header line
1481 contained an unqualified address, that address was qualified, even if it
1482 did not match any rewriting rules. The underlying bug was that the values
1483 of the flags that permit the existence of unqualified sender and recipient
1484 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1485 non-local messages, and by -bnq for local messages) were not being
1486 preserved with the message after it was received.
1488 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1489 "next input=" as part of the text comprising the host identity instead of
1490 the correct text. The code was using the same buffer for two different
1491 strings. However, depending on which order the printing function evaluated
1492 its arguments, the bug did not always show up. Under Linux, for example, my
1493 test suite worked just fine.
1495 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1496 doesn't work with some older versions of Perl. It has been changed to "my",
1497 which in any case is probably the better facility to use.
1499 15. A really picky compiler found some instances of statements for creating
1500 error messages that either had too many or two few arguments for the format
1503 16. The size of the buffer for calls to the DNS resolver has been increased
1504 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1505 for addresses that have a lot of PTR records. This alleviates a problem; it
1506 does not fully solve it.
1508 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1509 buffer now truncates the list and logs the incident, which is the same
1510 action as happens when Exim is looking up a host name and its aliases.
1511 Previously in this situation something unpredictable would happen;
1512 sometimes it was "internal error: store_reset failed".
1514 18. If a server dropped the connection unexpectedly when an Exim client was
1515 using GnuTLS and trying to read a response, the client delivery process
1516 crashed while trying to generate an error log message.
1518 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1519 string, for example:
1521 warn message = H1: something\nH2: something
1523 the text was added as a single header line from Exim's point of view
1524 though it ended up OK in the delivered message. However, searching for the
1525 second and subsequent header lines using $h_h2: did not work. This has been
1526 fixed. Similarly, if a system filter added multiple headers in this way,
1527 the routers could not see them.
1529 20. Expanded the error message when iplsearch is called with an invalid key to
1530 suggest using net-iplsearch in a host list.
1532 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1533 ACLs are no longer actually imposed (and a message to that effect is
1536 22. If a "gecos" field in a passwd entry contained escaped characters, in
1537 particular, if it contained a \" sequence, Exim got it wrong when building
1538 a From: or a Sender: header from that name. A second bug also caused
1539 incorrect handling when an unquoted " was present following a character
1540 that needed quoting.
1542 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1543 was not being matched caselessly.
1545 24. Arranged for all hyphens in the exim.8 source to be escaped with
1548 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1549 properly. Recipient callout cache records were still being keyed to include
1550 the sender, even when use_sender was set false. This led to far more
1551 callouts that were necessary. The sender is no longer included in the key
1552 when use_sender is false.
1554 26. Added "control = submission" modifier to ACLs.
1556 27. Added the ${base62d: operator to decode base 62 numbers.
1558 28. dnsdb lookups can now access SRV records.
1560 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1561 the configuration file.
1563 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1564 (-v) mode. This makes the output for a verbose queue run more intelligible.
1566 31. Added a use_postmaster feature to recipient callouts.
1568 32. Added the $body_zerocount variable, containing the number of binary zero
1569 bytes in the message body.
1571 33. The time of last modification of the "new" subdirectory is now used as the
1572 "mailbox time last read" when there is a quota error for a maildir
1575 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1577 35. Added +ignore_unknown as a special item in host lists.
1579 36. Code for decoding IPv6 addresses in host lists is now included, even if
1580 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1581 address was recognized as an IP address, but was then not correctly decoded
1582 into binary, causing unexpected and incorrect effects when compared with
1589 1. Very minor rewording of debugging text in manualroute to say "list of
1590 hosts" instead of "hostlist".
1592 2. If verify=header_syntax was set, and a header line with an unqualified
1593 address (no domain) and a large number of spaces between the end of the
1594 name and the colon was received, the reception process suffered a buffer
1595 overflow, and (when I tested it) crashed. This was caused by some obsolete
1596 code that should have been removed. The fix is to remove it!
1598 3. When running in the test harness, delay a bit after writing a bounce
1599 message to get a bit more predictability in the log output.
1601 4. Added a call to search_tidyup() just before forking a reception process. In
1602 theory, someone could use a lookup in the expansion of smtp_accept_max_
1603 per_host which, without the tidyup, could leave open a database connection.
1605 5. Added the variables $recipient_data and $sender_data which get set from a
1606 lookup success in an ACL "recipients" or "senders" condition, or a router
1607 "senders" option, similar to $domain_data and $local_part_data.
1609 6. Moved the writing of debug_print from before to after the "senders" test
1612 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1613 problems for message scanning, either using a data ACL, or using
1614 local_scan() because the Received: header was not generated till after they
1615 were called (in order to set the time as the time of reception completion).
1616 I have revised the way this works. The header is now generated after the
1617 body is received, but before the ACL or local_scan() are called. After they
1618 are run, the timestamp in the header is updated.
1624 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1625 before starting a queue runner without re-exec. This happened only when
1626 deliver_drop_privilege was set or when the Exim user was set to root. The
1627 effect of the bug was that timeouts during subsequent deliveries caused
1628 crashes instead of being properly handled. The handler is now left at its
1629 default (and expected) setting.
1631 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1632 message, again when deliver_drop_privilege is set or Exim is run as root.
1633 The bug described in (1) was not present in this case, but the tidying up
1634 of the other signals was missing. I have made the two cases consistent.
1636 3. The ignore_target_hosts setting on a manualroute router was being ignored
1637 for hosts that were looked up using the /MX notation.
1639 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1642 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1643 operated on the sender address. After changing the $sender_address to <>
1644 for the sender address verify, Exim was re-instated it as the original
1645 (before rewriting) address, but remembering that it had rewritten it, so it
1646 wasn't rewriting it again. This bug also had the effect of breaking the
1647 sender address verification caching when the sender address was rewritten.
1649 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1650 This has been changed so that if the ip literal address matches
1651 ignore_target_hosts, the router declines.
1653 7. Added expansion conditions match_domain, match_address, and match_local_
1654 part (NOT match_host).
1656 8. The placeholder for the Received: header didn't have a length field set.
1658 9. Added code to Exim itself and to exim_lock to test for a specific race
1659 condition that could lead to file corruption when using MBX delivery. The
1660 issue is with the lockfile that is created in /tmp. If this file is removed
1661 after a process has opened it but before that process has acquired a lock,
1662 there is the potential for a second process to recreate the file and also
1663 acquire a lock. This could lead to two Exim processes writing to the file
1664 at the same time. The added code performs the same test as UW imapd; it
1665 checks after acquiring the lock that its file descriptor still refers to
1666 the same named file.
1668 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1669 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1672 11. Added the smtp_active_hostname option. If used, this will typically be made
1673 to depend on the incoming interface address. Because $interface_address is
1674 not set up until the daemon has forked a reception process, error responses
1675 that can happen earlier (such as "too many connections") no longer contain
1678 12. If an expansion in a condition on a "warn" statement fails because a lookup
1679 defers, the "warn" statement is abandoned, and the next ACL statement is
1680 processed. Previously this caused the whole ACL to be aborted.
1682 13. Added the iplsearch lookup type.
1684 14. Added ident_timeout as a log selector.
1686 15. Added tls_certificate_verified as a log selector.
1688 16. Added a global option tls_require_ciphers (compare the smtp transport
1689 option of the same name). This controls incoming TLS connections.
1691 17. I finally figured out how to make tls_require_ciphers do a similar thing
1692 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1693 before starting the TLS session.
1695 18. Tabs are now shown as \t in -bP output.
1697 19. If the log selector return_path_on_delivery was set, Exim crashed when
1698 bouncing a message because it had too many Received: header lines.
1700 20. If two routers both had headers_remove settings, and the first one included
1701 a superfluous trailing colon, the final name in the first list and the
1702 first name in the second list were incorrectly joined into one item (with a
1703 colon in the middle).
1709 1. Added -C and -D options to the exinext utility, mainly to make it easier
1710 to include in the automated testing, but these could be helpful when
1711 multiple configurations are in use.
1713 2. The exinext utility was not formatting the output nicely when there was
1714 an alternate port involved in the retry record key, nor when there was a
1715 message id as well (for retries that were specific to a specific message
1716 and a specific host). It was also confused by IPv6 addresses, because of
1717 the additional colons they contain. I have fixed the IPv4 problem, and
1718 patched it up to do a reasonable job for IPv6.
1720 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1721 delivery, the log line now contains "pipelined" if PIPELINING was used.
1723 4. An SMTP transport process used to panic and die if the bind() call to set
1724 an explicit outgoing interface failed. This has been changed; it is now
1725 treated in the same way as a connect() failure.
1727 5. A reference to $sender_host_name in the part of a conditional expansion
1728 that was being skipped was still causing a DNS lookup. This no longer
1731 6. The def: expansion condition was not recognizing references to header lines
1732 that used bh_ and bheader_.
1734 7. Added the _cache feature to named lists.
1736 8. The code for checking quota_filecount in the appendfile transport was
1737 allowing one more file than it should have been.
1739 9. For compatibility with Sendmail, the command line option
1747 and sets the incoming protocol and host name (for trusted callers). The
1748 host name and its colon can be omitted when only the protocol is to be set.
1749 Note the Exim already has two private options, -pd and -ps, that refer to
1750 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1751 "s", but I don't think that's a major issue.
1753 10. A number of refactoring changes to the code, none of which should affect
1756 (a) The number of logging options was getting close to filling up the
1757 32-bit word that was used as a bit map. I have split them into two classes:
1758 those that are passed in the argument to log_write(), and those that are
1759 only ever tested independently outside of that function. These are now in
1760 separate 32-bit words, so there is plenty of room for expansion again.
1761 There is no change in the user interface or the logging behaviour.
1763 (b) When building, for example, log lines, the code previously used a
1764 macro that called string_cat() twice, in order to add two strings. This is
1765 not really sufficiently general. Furthermore, there was one instance where
1766 it was actually wrong because one of the argument was used twice, and in
1767 one call a function was used. (As it happened, calling the function twice
1768 did not affect the overall behaviour.) The macro has been replaced by a
1769 function that can join an arbitrary number of extra strings onto a growing
1772 (c) The code for expansion conditions now uses a table and a binary chop
1773 instead of a serial search (which was left over from when there were very
1774 few conditions). Also, it now recognizes conditions like "pam" even when
1775 the relevant support is not compiled in: a suitably worded error message is
1776 given if an attempt is made to use such a condition.
1778 11. Added ${time_interval:xxxxx}.
1780 12. A bug was causing one of the ddress fields not to be passed back correctly
1781 from remote delivery subprocesses. The field in question was not being
1782 subsequently used, so this caused to problems in practice.
1784 13. Added new log selectors queue_time and deliver_time.
1786 14. Might have fixed a bug in maildirsizefile handling that threw up
1787 "unexpected character" debug warnings, and recalculated the data
1788 unnecessarily. In any case, I expanded the warning message to give more
1791 15. Added the message "Restricted characters in address" to the statements in
1792 the default ACL that block characters like @ and % in local parts.
1794 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1795 Three changes have been made:
1797 (a) There was a serious bug; a negative response to MAIL caused the whole
1798 recipient domain to be cached as invalid, thereby blocking all messages
1799 to all local parts at the same domain, from all senders. This bug has
1800 been fixed. The domain is no longer cached after a negative response to
1801 MAIL if the sender used is not empty.
1803 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1806 (c) A new callout option, "use_sender" has been added for people who want
1807 the modified behaviour.
1813 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1816 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1817 indeed breaks things for older releases.
1819 3. Added additional logging to the case where there is a problem reading data
1820 from a filter that is running in a subprocess using a pipe, in order to
1821 try to track down a specific problem.
1823 4. Testing facility fudge: when running in the test harness and attempting
1824 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1825 getting "No route to host". Convert this to a timeout.
1827 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1830 6. Some OS don't have socklen_t but use size_t instead. This affects the
1831 fifth argument of getsockopt() amongst other things. This is now
1832 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1833 can be set for individual OS. I have set it for SunOS5, OSF1, and
1834 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1835 some earlier ones do not.
1837 7. Change 4.30/15 was not doing the test caselessly.
1839 8. The standard form for an IPv6 address literal was being rejected by address
1840 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1841 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1842 this, as well as the form without the "IPv6" on the front (but only when
1843 address literals are enabled, of course).
1845 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1847 10. Exim crashed if a message with an empty sender address specified by -f
1848 encountered a router with an errors_to setting. This could be provoked only
1849 by a command such as
1853 where an empty string was supplied; "<>" did not hit this bug.
1855 11. Installed PCRE release 4.5.
1857 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1858 remained set. It is now erased.
1860 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1861 times from message ids (which are base 36 rather than the normal 62).
1863 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1864 were being counted as actual protocol errors, and logged if the log
1865 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1866 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1867 and DATA following a set of rejected RCPTs do not count as protocol errors.
1868 In other words, Exim assumes they were pipelined, though this may not
1869 actually be the case. Of course, in all cases the client gets an
1870 appropriate error code.
1872 15. If a lookup fails in an ACL condition, a message about the failure may
1873 be available; it is used if testing the ACL cannot continue, because most
1874 such messages specify what the cause of the deferral is. However, some
1875 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1876 that caused an old message to be retained and used if a later statement
1877 caused a defer, replacing the real cause of the deferral.
1879 16. If an IP address had so many PTR records that the DNS lookup buffer
1880 was not large enough to hold them, Exim could crash while trying to process
1881 the truncated data. It now detects and logs this case.
1883 17. Further to 4.21/58, another change has been made: if (and only if) the
1884 first line of a message (the first header line) ends with CRLF, a bare LF
1885 in a subsequent header line has a space inserted after it, so as not to
1886 terminate the header.
1888 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1889 unnecessarily, used atoi() instead of strtol(), and didn't check the
1890 termination when getting file sizes from file names by regex.
1892 19. Completely re-implemented the support for maildirsize files, in the light
1893 of a number of problems with the previous contributed implementation
1894 (4.30/29). In particular:
1896 . If the quota is zero, the maildirsize file is maintained, but no quota is
1899 . If the maildir directory does not exist, it is created before any attempt
1900 to write a maildirsize file.
1902 . The quota value in the file is just a cache; if the quota is changed in
1903 the transport, the new value overrides.
1905 . A regular expression is available for excluding directories from the
1908 20. The autoreply transport checks the characters in options that define the
1909 message's headers; it allows continued headers, but it was checking with
1910 isspace() after an embedded newline instead of explicitly looking for a
1913 21. If all the "regular" hosts to which an address was routed had passed their
1914 expiry times, and had not reached their retry times, the address was
1915 bounced, even if fallback hosts were defined. Now Exim should go on to try
1918 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1919 equivalent code in the SMTP transport. Some hosts send humungous responses
1920 to HELO/EHLO, more than 1024 it seems.
1922 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1923 gives compiler warnings in some environments. I've now done it "properly",
1926 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1927 (because of problems with the built-in one) was declared to return uschar *
1928 instead of char *, causing compiler failure.
1930 25. Fixed a file descriptor leak when processing alias/forward files.
1932 26. Fixed a minor format string issue in dbfn.c.
1934 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1936 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1937 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1940 29. When a sender address is verified, it is cached, to save repeating the test
1941 when there is more than one recipient in a message. However, when the
1942 verification involves a callout, it is possible for different callout
1943 options to be set for different recipients. It is too complicated to keep
1944 track of this in the cache, so now Exim always runs a verification when a
1945 callout is required, relying on the callout cache for the optimization.
1946 The overhead is duplication of the address routing, but this should not be
1949 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
1950 to be verified with callout=postmaster, and the main callout worked but the
1951 postmaster check failed, the verification correctly failed. However, if a
1952 subsequent RCPT command asked for sender verification *without* the
1953 postmaster check, incorrect caching caused this verification also to fail,
1956 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1957 it was not caching the DNS options (qualify_single, search_parents) that
1958 were used when the lookup failed. A subsequent lookup with different
1959 options therefore always gave the same answer, though there were cases
1960 where it should not have. (Example: a "domains = !$mx_any" option on a
1961 dnslookup router: the "domains" option is always processed without any
1962 widening, but the router might have qualify_single set.) Now Exim uses the
1963 cached value only when the same options are set.
1965 32. Added John Jetmore's "exipick" utility to the distribution.
1967 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
1968 than a timeout (e.g. a certificate is required, and is not provided), an
1969 Exim server now closes the connection immediately. Previously it waited for
1970 the client to close - but if the client is SSL, it seems that they each
1971 wait for each other, leading to a delay before one of them times out.
1973 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1974 maintained 0.8.x compatibility because I don't think many are using it, and
1975 it is clearly obsolete.
1977 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1980 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1981 client certificate was expired. A simple patch fixes this, though I don't
1982 understand the full logic of why the verify callback is called multiple
1985 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1986 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1987 which causes problems with some clients (such as the Certicom SSL Plus
1988 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1989 disables the coutermeasure allowing Eudora to connect."
1991 38. Exim was not checking that a write() to a log file succeeded. This could
1992 lead to Bad Things if a log got too big, in particular if it hit a file
1993 size limit. Exim now panics and dies if it cannot write to a log file, just
1994 as it does if it cannot open a log file.
1996 39. Modified OS/Makefile-Linux so that it now contains
1998 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2000 The two -D definitions ensure that Exim is compiled with large file
2001 support, which makes it possible to handle log files that are bigger than
2004 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2005 instance) a domain was checked against a named list that involved a lookup,
2006 causing $domain_data to be set, then another domain was checked against the
2007 same list, then the first domain was re-checked, the value of $domain_data
2008 after the final check could be wrong. In particular, if the second check
2009 failed, it could be set empty. This bug probably also applied to
2012 41. The strip_trailing_dot option was not being applied to the address given
2013 with the -f command-line option.
2015 42. The code for reading a message's header from the spool was incrementing
2016 $received_count, but never initializing it. This meant that the value was
2017 incorrect (doubled) while delivering a message in the same process in which
2018 it was received. In the most common configuration of Exim, this never
2019 happens - a fresh exec is done - but it can happen when
2020 deliver_drop_privilege is set.
2022 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2023 it now includes up to 150 characters of the unexpected data in the log
2026 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2027 and building data strings. The size of both of these buffers was 10 000
2028 bytes - far larger than anybody would *ever* want, thought I. Needless to
2029 say, somebody hit the limit. I have increased the maximum line length to
2030 20 000 and the maximum data length of concatenated lines to 100 000. I have
2031 also fixed two bugs, because there was no checking on these buffers. Tsk,
2032 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2033 buffer is too small.
2035 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2036 lsearch lookups. Now it does.
2038 46. When parsing a route_list item in a manualroute router, a fixed-length
2039 buffer was used for the list of hosts. I made this 1024 bytes long,
2040 thinking that nobody would ever have a list of hosts that long. Wrong.
2041 Somebody had a whole pile of complicated expansion conditions, and the
2042 string was silently truncated, leading to an expansion error. It turns out
2043 that it is easier to change to an unlimited length (owing to other changes
2044 that have happened since this code was originally written) than to build
2045 structure for giving a limitation error. The length of the item that
2046 expands into the list of hosts is now unlimited.
2048 47. The lsearch lookup could not handle data where the length of text line was
2049 more than 4095 characters. Such lines were truncated, leading to shortened
2050 data being returned. It should now handle lines of any length.
2052 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2053 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2056 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2057 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2058 no longer happens in comment lines. A list of replacements is now placed
2059 at the head of all of the source files, except those whose only change is
2060 to replace PERL_COMMAND in the very first #! line.
2062 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2063 messages on the queue, with a bottom-up merge sort, using code contributed
2064 by Michael Haardt. This should make operations like -bp somewhat faster on
2065 large queues. It won't affect queue runners, except when queue_run_in_order
2068 51. Installed eximstats 1.31 in the distribution.
2070 52. Added support for SRV lookups to the dnslookup router.
2072 53. If an ACL referred to $message_body or $message_body_end, the value was not
2073 reset for any messages that followed in the same SMTP session.
2075 54. The store-handling optimization for building very long strings was not
2076 differentiating between the different store pools. I don't think this
2077 actually made any difference in practice, but I've tidied it.
2079 55. While running the routers to verify a sender address, $sender_address
2080 was still set to the sender address. This is wrong, because when routing to
2081 send a bounce to the sender, it would be empty. Therefore, I have changed
2082 it so that, while verifying a sender address, $sender_address is set to <>.
2083 (There is no change to what happens when verifying a recipient address.)
2085 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2086 target A or AAAA records (if not already returned) without resetting the
2087 qualify_single or search_parents options of the DNS resolver. These are
2088 inappropriate in this case because the targets of MX and SRV records must
2089 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2090 target that, when qualified, matched something in the local domain. These
2091 two options are now turned off when doing these lookups.
2093 57. It seems that at least some releases of Reiserfs (which does not have the
2094 concept of a fixed number of inodes) returns zero and not -1 for the
2095 number of available inodes. This interacted badly with check_spool_inodes,
2096 which assumed that -1 was the "no such thing" setting. What I have done is
2097 to check that the total number of inodes is greater than zero before doing
2098 the test of how many are available.
2100 58. When a "warn" ACL statement has a log_message modifier, the message is
2101 remembered, and not repeated. This is to avoid a lot of repetition when a
2102 message has many recipients that cause the same warning to be written.
2103 Howewer, Exim was preserving the list of already written lines for an
2104 entire SMTP session, which doesn't seem right. The memory is now reset if a
2105 new message is started.
2107 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2108 debugging output unless log_rewrite was also set.
2110 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2111 of (int)(handle) when we know that handle contains (void *)(-1).
2113 61. The Exim daemon panic-logs an error return when it closes the incoming
2114 connection. However "connection reset by peer" seems to be common, and
2115 isn't really an error worthy of noting specially, so that particular error
2118 62. When Exim is trying to find all the local interfaces, it used to panic and
2119 die if the ioctl to get the interface flags failed. However, it seems that
2120 on at least one OS (Solaris 9) it is possible to have an interface that is
2121 included in the list of interfaces, but for which you get a failure error
2122 for this call. This happens when the interface is not "plumbed" into a
2123 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2124 failure of the "get flags" call assumes that the interface is down.
2126 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2127 makes life easier for people who are doing arithmetic on fields extracted
2128 from dates, where you often get leading zeros that should not be
2129 interpreted as octal.
2131 64. Added qualify_domain to the redirect router, to override the global
2134 65. If a pathologically long header line contained very many addresses (the
2135 report of this problem mentioned 10 000) and each of them was rewritten,
2136 Exim could use up a very large amount of memory. (It kept on making new
2137 copies of the header line as it rewrote, and never released the old ones.)
2138 At the expense of a bit more processing, the header rewriting function has
2139 been changed so that it no longer eats memory in this way.
2141 66. The generation of the Received: header has been moved from the time that a
2142 message starts to be received, to the time that it finishes. The timestamp
2143 in the Received: header should now be very close to that of the <= log
2144 line. There are two side-effects of this change:
2146 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2147 logged header lines no longer include the local Received: line, because
2148 it has not yet been created. The same applies to a copy of the message
2149 that is returned to a non-SMTP sender when a message is rejected.
2151 (b) When a filter file is tested using -bf, no additional Received: header
2152 is added to the test message. After some thought, I decided that this
2155 This change does not affect the value of $received_for. It is still set
2156 after address rewriting, but before local_scan() is called.
2158 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2160 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2161 gave an unhelpful panic error message, and a defer error. I have managed to
2162 change this behaviour so that it now rejects any supplied certificate,
2163 which seems right, as the list of acceptable certificates is empty.
2165 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2166 gave an unhelpful defer error. I have not managed to make this reject any
2167 supplied certificates, but the error message it gives is "no certificate
2168 supplied", which is not helpful.
2170 70. exigrep's output now also includes lines that are not associated with any
2171 message, but which match the given pattern. Implemented by a patch from
2172 Martin Sluka, which also tidied up the Perl a bit.
2174 71. Recipient callout verification, like sender verification, was using <> in
2175 the MAIL FROM command. This isn't really the right thing, since the actual
2176 sender may affect whether the remote host accepts the recipient or not. I
2177 have changed it to use the actual sender in the callout; this means that
2178 the cache record is now keyed on a recipient/sender pair, not just the
2179 recipient address. There doesn't seem to be a real danger of callout loops,
2180 since a callout by the remote host to check the sender would use <>.
2181 [SEE ABOVE: changed after hitting problems.]
2183 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2184 temporary errors. However, in the case of such a code being given after
2185 the end of a data transmission (i.e. after ".") Exim was failing to write
2186 a retry record for the message. (Yes, there was some broken host that was
2187 actually sending 8xx at this point.)
2189 73. An unknown lookup type in a host list could cause Exim to panic-die when
2190 the list was checked. (An example that provoked this was putting <; in the
2191 middle of a list instead of at the start.) If this happened during a DATA
2192 ACL check, a -D file could be left lying around. This kind of configuration
2193 error no longer causes Exim to die; instead it causes a defer errror. The
2194 incident is still logged to the main and panic logs.
2196 74. Buglet left over from Exim 3 conversion. The message "too many messages
2197 in one connection" was written to the rejectlog but not the mainlog, except
2198 when address rewriting (yes!) was being logged.
2200 75. Added write_rejectlog option.
2202 76. When a system filter was run not as root (that is, when system_filter_user
2203 was set), the values of the $n variables were not being returned to the
2204 main process; thus, they were not subsequently available in the $sn
2207 77. Added +return_path_on_delivery log selector.
2209 78. A connection timeout was being treated differently from recipients deferred
2210 when testing hosts_max_try with a message that was older than the host's
2211 retry timeout. (The host should not be counted, thus allowing all hosts to
2212 be tried at least once before bouncing.) This may have been the cause of an
2213 occasionally reported bug whereby a message would remain on the queue
2214 longer than the retry timeout, but would be bounced if a delivery was
2215 forced. I say "may" because I never totally pinned down the problem;
2216 setting up timeout/retry tests is difficult. See also the next item.
2218 79. The ultimate address timeout was not being applied to errors that involved
2219 a combination of host plus message (for example, a timeout on a MAIL
2220 command). When an address resolved to a number of possible hosts, and they
2221 were not all tried for each delivery (e.g. because of hosts_max_try), a
2222 message could remain on the queue longer than the retry timeout.
2224 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2227 81. Fixed an obscure SMTP outgoing bug which required at least the following
2228 conditions: (a) there was another message waiting for the same server;
2229 (b) the server returned 5xx to all RCPT commands in the first message so
2230 that the message was not completed; (c) the server dropped the connection
2231 or gave a negative response to the RSET that Exim sends to abort the
2232 transaction. The observed case was a dropped connection after DATA that had
2233 been sent in pipelining mode. That is, the server had advertised PIPELINING
2234 but was not implementing it correctly. The effect of the bug was incorrect
2235 behaviour, such as trying another host, and this could lead to a crash.
2241 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2242 and daemon.c were passed as pointers to ints; they should have been
2243 pointers to socklen_t variables (which are typically unsigned ints).
2245 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2248 3. Fixed a really odd bug that affected only the testing scheme; patching a
2249 certain fixed string in the binary changed the value of another string that
2250 happened to be identical to the end of the original first string.
2252 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2253 name", it returns that address as the IP address. On some operating
2254 systems (e.g. Solaris), it also passes back the IP address string as the
2255 "host name". However, on others (e.g. Linux), it passes back an empty
2256 string. Exim wasn't checking for this, and was changing the host name to an
2257 empty string, assuming it had been canonicized.
2259 5. Although rare, it is permitted to have more than one PTR record for a given
2260 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2261 all the names associated with an address, because they do in Solaris.
2262 However, it seems that they do not in Linux for data that comes from the
2263 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2264 I found this out when I moved to a new Linux workstation and tried to run
2265 the Exim test suite.
2267 To get round this problem I have changed the code so that it now does its
2268 own call to the DNS to look up PTR records when searching for a host name.
2269 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2270 addresses that are only in /etc/hosts are still found.
2272 This behaviour is, however, controlled by an option called host_lookup_
2273 order, which defaults to "bydns:byaddr". If people want to use the other
2274 order, or indeed, just use one or the other means of lookup, they can
2275 specify it in this variable.
2277 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2278 some operating systems, this comes back from gethostbyaddr() as an empty
2279 string, and this is what Exim used to test for. However, it seems that in
2280 other systems, "." is yielded. Exim now tests for this case too.
2282 7. The values of check_spool_space and check_log_space are now held internally
2283 as a number of kilobytes instead of an absolute number of bytes. If a
2284 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2285 kilobyte. This means that much larger values can be stored.
2287 8. Exim monitor: an attempt to get the action menu when not actually pointing
2288 at a message produces an empty menu entitled "No message selected". This
2289 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2290 no entries in it ("Shell widget menu has zero width and/or height"). So I
2291 have added a single, blank menu entry in this case.
2293 9. Added ${quote_local_part.
2295 10. MIME decoding is now applied to the contents of Subject: header lines when
2298 11. Now that a reference to $sender_host_address automatically causes a reverse
2299 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2300 host lookup before query-style lookups in lists that might use this
2301 variable. This has therefore been abolished, and the "net-" prefix is no
2302 longer necessary for query-style lookups.
2304 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2305 have been a typo for LFLAGS, so it has been changed.
2307 13. The install script calls Exim with "-C /dev/null" in order to find the
2308 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2309 to be output. Howeve, since Exim outputs its version number before the
2310 error, it didn't break the script. It just looked ugly. I fixed this by
2311 always allowing "-C /dev/null" if the caller is root.
2313 14. Ignore overlarge ACL variable number when reading spool file - insurance
2314 against a later release with more variables having written the file.
2316 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2317 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2318 this, as well as the form without the "IPv6" on the front.
2320 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2321 OS/Makefile-Darwin file.
2323 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2324 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2326 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2327 variable. However, some OS use ints for inodes. I've added cast to long int
2328 to get rid of the compiler warning.
2330 19. I had forgotten to lock out "/../" in configuration file names when
2331 ALT_CONFIG_PREFIX was set.
2333 20. Routers used for verification do not need to specify transports. However,
2334 if such a router generated a host list, and callout was configured, Exim
2335 crashed, because it could not find a port number from the (non-existent)
2336 transport. It now assumes port 25 in this circumstance.
2338 21. Added the -t option to exigrep.
2340 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2341 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2342 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2343 them is set, LOOKUP_LSEARCH is forced.
2345 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2347 24. Added sender and host information to the "rejected by local_scan()" log
2348 line; previously there was no indication of these.
2350 25. Added .include_if_exists.
2352 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2353 Linux. It turns out that not all file systems support this. Apparently some
2354 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2355 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2356 not supported on the file descriptor, is now ignored when Exim is trying to
2357 sync a directory. This applies only to Linux.
2359 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2361 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2362 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2365 29. Added support for maildirsize files from supplied patch (modified a bit).
2367 30. The use of :fail: followed by an empty string could lead Exim to respond to
2368 sender verification failures with (e.g.):
2370 550 Verification failed for <xxx>
2371 550 Sender verify failed
2373 where the first response line was missing the '-' that indicates it is not
2374 the final line of the response.
2376 31. The loop for finding the name of the user that called Exim had a hardwired
2377 limit of 10; it now uses the value of finduser_retries, which is used for
2378 all other user lookups.
2380 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2383 33. Exim was neglecting to zero errno before one call of strtol() when
2384 expanding a string and expecting an integer value. On some systems this
2385 resulted in spurious "integer overflow" errors. Also, it was casting the
2386 result into an int without checking.
2388 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2389 did not work. The code looks as if it has *never* worked, though it appears
2390 to have been documented since at least releast 1.62. I have made it work.
2392 35. The "timeout_DNS" error in retry rules, also documented since at least
2393 1.62, also never worked. As it isn't clear exactly what this means, and
2394 clearly it isn't a major issue, I have abolished the feature by treating it
2395 as "timeout", and writing a warning to the main and panic logs.
2397 36. The display of retry rules for -brt wasn't always showing the error code
2400 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2401 timeout_connect_A, timeout_connect_MX.
2403 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2404 to the empty sender.
2406 39. The daemon was not analysing the content of -oX till after it had closed
2407 stderr and disconnected from the controlling terminal. This meant that any
2408 syntax errors were only noted on the panic log, and the return code from
2409 the command was 0. By re-arranging the code a little, I've made the
2410 decoding happen first, so such errors now appear on stderr, and the return
2411 code is 1. However, the actual setting up of the sockets still happens in
2412 the disconnected process, so errors there are still only recorded on the
2415 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2416 connections (as happens on some IP stacks) was logged at start up time as
2417 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2418 it from "IPv6 and IPv4", which means that two separate sockets are being
2421 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2422 says whether AF_INET or AF_INET6 was passed as an argument.
2424 42. Exiwhat output was messed up when time zones were included in log
2427 43. Exiwhat now gives more information about the daemon's listening ports,
2428 and whether -tls-on-connect was used.
2430 44. The "port" option of the smtp transport is now expanded.
2432 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2433 silently ignored. Now an error message is written to the main and panic
2436 46. There's a new ACL modifier called "logwrite" which writes to a log file
2437 as soon as it is encountered.
2439 47. Added $local_user_uid and $local_user_gid at routing time.
2441 48. Exim crashed when trying to verify a sender address that was being
2444 49. Exim was recognizing only a space character after ".include". It now also
2445 recognizes a tab character.
2447 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2448 extracting the relevant information from the specification. The man page no
2449 longer contains scrambled data for the -d option, and I've added a section
2450 at the front about calling Exim under different names.
2452 51. Added "extra_headers" argument to the "mail" command in filter files.
2454 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2457 53. Installed eximstats 1.29.
2459 54. Added transport_filter_timeout as a generic transport option.
2461 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2462 Cc: header lines. This was required by RFC 822, but it not required by RFC
2465 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2466 incoming messages that did not have them. Now it does so only if the
2467 message originates locally, that is, if there is no associated remote host
2468 address. When Resent- header lines are present, this applies to the Resent-
2469 lines rather than the non-Resent- lines.
2471 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2472 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2474 58. Messages for configuration errors now include the name of the main
2475 configuration file - useful now that there may be more than one file in a
2476 list (.included file names were always shown).
2478 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2479 for those rare installations that do not start the daemon as root or run it
2480 setuid root. I've cut out the call to initgroups() if the daemon is not
2483 60. The Exim user and group can now be bound into the binary as text strings
2484 that are looked up at the start of Exim's processing.
2486 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2488 62. Added $mailstore_basename variable.
2490 63. Installed patch to sieve.c from Michael Haardt.
2492 64. When Exim failed to open the panic log after failing to open the main log,
2493 the original message it was trying to log was written to stderr and debug
2494 output, but if they were not available (the usual case in production), it
2495 was lost. Now it is written to syslog before the two lines that record the
2496 failures to open the logs.
2498 65. Users' Exim filters run in subprocesses under the user's uid. It is
2499 possible for a "deliver" command or an alias in a "personal" command to
2500 provoke an address rewrite. If logging of address rewriting is configured,
2501 this fails because the process is not running as root or exim. There may be
2502 a better way of dealing with this, but for the moment (because 4.30 needs
2503 to be released), I have disabled address rewrite logging when running a
2504 filter in a non-root, non-exim process.
2510 1. The buildconfig auxiliary program wasn't quoting the value set for
2511 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2512 not defined. This bug was masked in 4.22 by the effect that was fixed in
2515 2. Some messages that were rejected after a message id was allocated were
2516 shown as "incomplete" by exigrep. It no longer does this for messages that
2517 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2519 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2520 have allow_domain_literals set, the ID did not get logged in the <= line.
2521 Domain literals are now always recognized in Message-ID: header lines.
2523 4. The first argument for a ${extract expansion item is the key name or field
2524 number. Leading and trailing spaces in this item were not being ignored,
2525 causing some misleading effects.
2527 5. When deliver_drop_privilege was set, single queue runner processes started
2528 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2529 same command in the process it spins off) were not dropping privilege.
2531 6. When the daemon running as "exim" started a queue runner, it always
2532 re-executed Exim in the spun-off process. This is a waste of effort when
2533 deliver_drop_privilege is set. The new process now just calls the
2534 queue-runner function directly.
2540 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2543 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2544 ignored. Though the use of -oP was forcing the writing of a pid file, it
2545 was always written to the default place.
2547 3. If the message "no IP address found for host xxxx" is generated during
2548 incoming verification, it is now followed by identification of the incoming
2549 connection (so you can more easily find what provoked it).
2551 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2553 5. Added some features to "harden" Exim a bit more against certain attacks:
2555 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2556 be put in Local/Makefile. This is like the never_users runtime option,
2557 but it cannot be overridden. The default setting is "root".
2559 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2560 prefix string with which any file named in a -C command line option
2563 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2564 is retained for -C and -D only if the caller of Exim is root. Without
2565 it, the exim user may also use -C and -D and retain privilege.
2567 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2568 command line option is disabled.
2570 6. Macro names set by the -D option must start with an upper case letter, just
2571 like macro names defined in the configuration file.
2573 7. Added "dereference=" facility to LDAP.
2575 8. Two instances of the typo "uknown" in the source files are fixed.
2577 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2578 the Configure-Makefile script screwed up while processing it.
2580 10. Incorporated PCRE 4.4.
2582 11. The SMTP synchronization check was not operating right at the start of an
2583 SMTP session. For example, it could not catch a HELO sent before the client
2584 waited for the greeting. There is now a check for outstanding input at the
2585 point when the greeting is written. Because of the duplex, asynchronous
2586 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2587 way, but not yet received, when the check is performed.
2589 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2590 on TCP/IP sockets, because this apparently causes some broken clients to
2593 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2594 unchanged) from the Cygwin maintainer.
2596 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2597 was supported instead of testing for mbx. Effectively a typo.
2599 15. The spa authenticator server code was not checking that the input it
2600 received was valid base64.
2602 16. The debug output line for the "set" modifier in ACLs was not showing the
2603 name of the variable that was being set.
2605 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2607 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2608 lookup on its own. Something else was needed to trigger the lookup. For
2609 example, a match in host_lookup or the need for a host name in a host list.
2610 Now, if $sender_host_name is referenced and the host name has not yet been
2611 looked up, a lookup is performed. If the lookup fails, the variable remains
2612 empty, and $host_lookup_failed is set to "1".
2614 19. Added "eqi" as a case-independent comparison operator.
2616 20. The saslauthd authentication condition could segfault if neither service
2617 nor realm was specified.
2619 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2620 error message that was logged was misleading, and incoming SMTP
2621 connections were dropped. The message is now more accurate, and temporary
2622 errors are given to SMTP connections.
2624 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2625 (except RSET) with a 421 error, until QUIT is received. However, it was
2626 failing to send a response to QUIT.
2628 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2629 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2630 ACL is now run after the helo_try_verify_hosts code.
2632 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2633 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2634 case-independent, so other case variants are also recognized). Apparently
2635 some systems use these upper case variants.
2637 25. If more than two messages were waiting for the same host, and a transport
2638 filter was specified for the transport, Exim sent two messages over the
2639 same TCP/IP connection, and then failed with "socket operation on non-
2640 socket" when it tried to send the third.
2642 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2644 27. The extern definition of crypt16() in expand.c was not being excluded when
2645 the OS had its own crypt16() function.
2647 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2648 as a preferred synonym for return_size_limit, both as an option and as an
2651 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2653 30. Changed the default configuration ACL to relax the local part checking rule
2654 for addresses that are not in any local domains. For these addresses,
2655 slashes and pipe symbols are allowed within local parts, but the sequence
2656 /../ is explicitly forbidden.
2658 31. SPA server authentication was not clearing the challenge buffer before
2661 32. log_message in a "warn" ACL statement was writing to the reject log as
2662 well as to the main log, which contradicts the documentation and doesn't
2663 seem right (because no rejection is happening). So I have stopped it.
2665 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2666 However, I am unable to do any testing of this.
2668 34. Fixed an infelicity in the appendfile transport. When checking directories
2669 for a mailbox, to see if any needed to be created, it was accidentally
2670 using path names with one or more superfluous leading slashes; tracing
2671 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2673 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2674 contents are added to the log line that is written for every discarded
2675 recipient. (Previously a log_message setting was ignored.)
2677 36. The ${quote: operator now quotes the string if it is empty.
2679 37. The install script runs exim in order to find its version number. If for
2680 some reason other than non-existence or emptiness, which it checks, it
2681 could not run './exim', it was installing it with an empty version number,
2682 i.e. as "exim-". This error state is now caught, and the installation is
2685 38. An argument was missing from the function that creates an error message
2686 when Exim fails to connect to the socket for saslauthd authentication.
2687 This could cause Exim to crash, or give a corrupted message.
2689 39. Added isip, isip4, and isip6 to ${if conditions.
2691 40. The ACL variables $acl_xx are now saved with the message, and can be
2692 accessed later in routers, transports, and filters.
2694 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2695 strings in the file are not string-expanded.
2697 42. If a MAIL command specified a SIZE value that was too large to fit into an
2698 int variable, the check against message_size_limit failed. Such values are
2699 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2700 day this will have to be increased, but I don't think I want to be around
2701 when emails are that large.
2708 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2709 iconv() is not standard in FreeBSD.
2711 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2712 IPv6 enabled. The observed symptom was a segmentation fault on return from
2713 the function os_common_find_running_interfaces() in src/os.c.
2715 3. In the check_special_case() function in daemon.c I had used "errno" as an
2716 argument name, which causes warnings on some systems. This was basically a
2717 typo, since it was named "eno" in the comments!
2719 4. The code that waits for the clock to tick (at a resolution of some fraction
2720 of a second) so as to ensure message-id uniqueness was always waiting for
2721 at least one whole tick, when it could have waited for less. [This is
2722 almost certainly not relevant at current processor speeds, where it is
2723 unlikely to ever wait at all. But we try to future-proof.]
2725 5. The function that sleeps for a time interval that includes fractions of a
2726 second contained a race. It did not block SIGALRM between setting the
2727 timer, and suspending (a couple of lines later). If the interval was short
2728 and the sigsuspend() was delayed until after it had expired, the suspension
2729 never ended. On busy systems this could lead to processes getting stuck for
2732 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2733 process, before it forks any delivery processes. The open lookup caching
2734 mechanism meant that the open file or database connection was passed into
2735 the delivery process. The problem was that delivery processes always tidy
2736 up cached lookup data. This could cause a problem for the next delivery
2737 process started by the queue runner, because the external queue runner
2738 process does not know about the closure. So the next delivery process
2739 still has data in the lookup cache. In the case of a file lookup, there was
2740 no problem because closing a file descriptor in a subprocess doesn't affect
2741 the parent. However, if the lookup was caching a connection to a database,
2742 the connection was closed, and the second delivery process was likely to
2743 see errors such as "PGSQL: query failed: server closed the connection
2744 unexpectedly". The problem has been fixed by closing all cached lookups
2745 in a queue runner before running a delivery process.
2747 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2748 seem to have the "const" qualifier which it has on other OS. I've
2751 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2752 *of the same type* (client or server) with the same public name that an
2753 error should be diagnosed.
2755 9. When Exim looked up a host name for an IP address, but failed to find the
2756 original IP address when looking up the host name (a safety check), it
2757 output the message "<ip address> does not match any IP for NULL", which was
2758 confusing, to say the least. The bug was that the host name should have
2759 appeared instead of "NULL".
2761 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2762 user that is built into the binary, and the -C or -D options is used, root
2763 privilege is dropped before the configuration file is read. In addition,
2764 logging is switched to stderr instead of the normal log files. If the
2765 configuration then re-defines the Exim user, the unprivileged environment
2766 is probably not what is expected, so Exim logs a panic warning message (but
2769 However, if deliver_drop_privilege is set, the unprivileged state may well
2770 be exactly what is intended, so the warning has been cut out in that case,
2771 and Exim is allowed to try to write to its normal log files.
2777 1. smtp_return_error_details was not giving details for temporary sender
2778 or receiver verification errors.
2780 2. Diagnose a configuration error if two authenticators have the same public
2783 3. Exim used not to create the message log file for a message until the first
2784 delivery attempt. This could be confusing when incoming messages were held
2785 for policy or load reasons. The message log file is now created at the time
2786 the message is received, and an initial "Received" line is written to it.
2788 4. The automatically generated man page for command line options had a minor
2789 bug that caused no ill effects; however, a more serious problem was that
2790 the procedure for building the man page automatically didn't always
2791 operate. Consequently, release 4.20 contains an out-of-date version. This
2792 shouldn't happen again.
2794 5. When building Exim with embedded Perl support, the script that builds the
2795 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2796 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2798 6. The freeze_tell option was not being used for messages that were frozen on
2799 arrival, either by an ACL or by local_scan().
2801 7. Added the smtp_incomplete_transaction log selector.
2803 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2804 was accepting AUTH without a new EHLO.
2806 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2807 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2810 10. Logging of TCP/IP connections (when configured) now happens in the main
2811 daemon process instead of the child process, so that the TCP/IP connection
2812 count is more accurate (but it can never be perfect).
2814 11. The use of "drop" in a nested ACL was not being handled correctly in the
2815 outer ACL. Now, if condition failure induced by the nested "drop" causes
2816 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2817 or "require"), the connection is dropped.
2819 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2820 that yield "discard" can now be used with an "accept" or a "discard" verb,
2821 but an error is generated for any others (because I can't see a useful way
2822 to define what should happen).
2824 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2825 are now processed in the same way as lines in the Exim configuration file.
2826 In particular, continuation lines are supported.
2828 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2830 15. Added -ti meaning -t -i.
2832 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2833 domains, and warn by logging if others are found.
2835 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2836 returned from GIFCONF when Exim is trying to find the list of interfaces on
2837 a host. The code in os.c has been modified to copy each ifreq to an aligned
2838 structure in all cases.
2840 Also, in some cases, the returned ifreq's were being copied to a 'struct
2841 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2842 means the last couple of bytes of an IPv6 address could be chopped if the
2843 ifreq contained only a normal sockaddr (14 bytes storage).
2845 18. Named domain lists were not supported in the hosts_treat_as_local option.
2846 An entry such as +xxxx was not recognized, and was treated as a literal
2849 19. Ensure that header lines added by a DATA ACL are included in the reject log
2850 if the ACL subsequently rejects the message.
2852 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2853 MD5 (which is deprecated).
2855 21. When testing a filter file using -bf, Exim was writing a message when it
2856 took the sender from a "From " line in the message, but it was not doing so
2857 when it took $return_path from a Return-Path: header line. It now does.
2859 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2860 with a valid header line field name (a series of printing characters
2861 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2863 23. Changed "disc" in the source to "disk" to conform to the documentation and
2864 the book and for uniformity.
2866 24. Ignore Sendmail's -Ooption=value command line item.
2868 25. When execve() failed while trying to run a command in a pipe transport,
2869 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2870 could be confused with a return value of 69 from the command itself. This
2871 has been changed to 127, the value the shell returns if it is asked to run
2872 a non-existent command. The wording for the related log line suggests a
2873 non-existent command as the problem.
2875 26. If received_header_text expands to an empty string, do not add a Received:
2876 header line to the message. (Well, it adds a token one on the spool, but
2877 marks it "old" so that it doesn't get used or transmitted.)
2879 27. Installed eximstats 1.28 (addition of -nt option).
2881 28. There was no check for failure on the call to getsockname() in the daemon
2882 code. This can fail if there is a shortage of resources on the system, with
2883 ENOMEM, for example. A temporary error is now given on failure.
2885 29. Contrary to the C standard, it seems that in some environments, the
2886 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2887 program. Exim now does this explicitly; it affects the formatting of
2888 timestamps using strftime().
2890 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2891 complaints. I've now initialized all the variables, to avoid this.
2893 32. Header lines added by a system filter were not being "seen" during
2894 transport-time rewrites.
2896 33. The info_callback() function passed to OpenSSL is set up with type void
2897 (*)(SSL *, int, int), as described somewhere. However, when calling the
2898 function (actually a macro) that sets it up, the type void(*)() is
2899 expected. I've put in a cast to prevent warnings from picky compilers.
2901 34. If a DNS black list lookup found a CNAME record, but there were no A
2902 records associated with the domain it pointed at, Exim crashed.
2904 35. If a DNS black list lookup returned more than one A record, Exim ignored
2905 all but the first. It now scans all returned addresses if a particular IP
2906 value is being sought. In this situation, the contents of the
2907 $dnslist_value variable are a list of all the addresses, separated by a
2910 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2911 to accept a host name and all its aliases if the forward lookup for any of
2912 them yielded the IP address of the incoming connection. Now it accepts only
2913 those names whose forward lookup yields the correct IP address. Any other
2914 names are discarded. This closes a loophole whereby a rogue DNS
2915 administrator could create reverse DNS records to break through a
2916 wildcarded host restriction in an ACL.
2918 37. If a user filter or a system filter that ran in a subprocess used any of
2919 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2920 the wrong values were passed to the pipe command ($thisaddress had the
2921 value of $0, $0 had the value of $1, etc). This bug was introduced by
2922 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2924 38. Improved the line breaking for long SMTP error messages from ACLs.
2925 Previously, if there was no break point between 40 and 75 characters, Exim
2926 left the rest of the message alone. Two changes have been made: (a) I've
2927 reduced the minimum length to 35 characters; (b) if it can't find a break
2928 point between 35 and 75 characters, it looks ahead and uses the first one
2929 that it finds. This may give the occasional overlong line, but at least the
2930 remaining text gets split now.
2932 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2933 file descriptors might be low, and that setting 1000 would always raise it.
2934 It turns out that in some environments, the limit is already over 1000 and
2935 that lowering it causes trouble. So now Exim takes care not to decrease it.
2937 40. When delivering a message, the value of $return_path is set to $sender_
2938 address at the start of routing (routers may change the value). By an
2939 oversight, this default was not being set up when an address was tested by
2940 -bt or -bv, which affected the outcome if any router or filter referred to
2943 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2944 log only when "message" or "log_message" are set. However, if one of the
2945 conditions was an address verification, or a call to a nested ACL, the
2946 messages generated by the underlying test were being passed through. This
2947 no longer happens. The underlying message is available in $acl_verify_
2948 message for both "message" and "log_message" expansions, so it can be
2949 passed through if needed.
2951 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2952 new expansion $bh_ to give the encoded byte string without charset
2953 translation. Translation happens only if iconv() is available; HAVE_ICONV
2954 indicates this at build time. HEADERS_CHARSET gives the charset to
2955 translate to; headers_charset can change it in the configuration, and
2956 "headers charset" can change it in an individual filter file.
2958 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
2959 that creates RFC 2047 encoded "words" labels them as that charset instead
2960 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2961 expansion operator; (ii) when Exim creates a From: line for a local
2962 message; (iii) when a header line is rewritten to include a "phrase" part.
2964 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2965 buggy, causing it to skip the first lines of messages whose message ID
2966 ended in 'D'. This would not have bitten before Exim release 4.14, because
2967 message IDs were unlikely to end in 'D' before then. The effect was to have
2968 incorrect size information for certain domains.
2970 45. #include "config.h" was missing at the start of the crypt16.c module. This
2971 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2974 46. If there was a timeout during a "random" callout check, Exim treated it as
2975 a failure of the random address, and carried on sending RSET and the real
2976 address. If the delay was just some slowness somewhere, the response to the
2977 original RCPT would be taken as a response to RSET and so on, causing
2978 mayhem of various kinds.
2980 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2981 when I implemented it. It didn't allow for the fact that some option values
2982 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2983 the right test for positive values.
2985 48. Domain names in DNS records are case-independent. Exim always looks them up
2986 in lower case. Some resolvers return domain names in exactly the case they
2987 appear in the zone file, that is, they may contain uppercase letters. Not
2988 all resolvers do this - some return always lower case. Exim was treating a
2989 change of case by a resolver as a change of domain, similar to a widening
2990 of a domain abbreviation. This triggered its re-routing code and so it was
2991 trying to route what was effectively the same domain again. This normally
2992 caused routing to fail (because the router wouldn't handle the domain
2993 twice). Now Exim checks for this case specially, and just changes the
2994 casing of the domain that it ultimately uses when it transmits the message
2997 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3000 50. If a filter generated a file delivery with a non-absolute name (possible if
3001 no home directory exists for the router), the forbid_file option was not
3004 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3005 the existing equality matching.
3007 52. Exim was using ints instead of ino_t variables in some places where it was
3008 dealing with inode numbers.
3010 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3011 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3012 called TMPDIR, and if it finds it is different, it changes its value.
3014 54. The smtp_printf() function is now made available to local_scan() so
3015 additional output lines can be written before returning. There is also an
3016 smtp_fflush() function to enable the detection of a dropped connection.
3017 The variables smtp_input and smtp_batched_input are exported to
3020 55. Changed the default runtime configuration: the message "Unknown user"
3021 has been removed from the ACL, and instead placed on the localuser router,
3022 using the cannot_route_message feature. This means that any verification
3023 failures that generate their own messages won't get overridden. Similarly,
3024 the "Unrouteable address" message that was in the ACL for unverifiable
3025 relay addresses has also been removed.
3027 56. Added hosts_avoid_esmtp to the smtp transport.
3029 57. The exicyclog script was not checking for the esoteric option
3030 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3031 will work only if exicyclog is run under the appropriate euid.
3033 58. Following a discussion on the list, the rules by which Exim recognises line
3034 endings on incoming messages have been changed. The -dropcr and drop_cr
3035 options are now no-ops, retained only for backwards compatibility. The
3036 following line terminators are recognized: LF CRLF CR. However, special
3037 processing applies to CR:
3039 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3040 nor a local message in the state where . is a terminator.
3042 (ii) If a bare CR is encountered in a header line, an extra space is added
3043 after the line terminator so as not to end the header. The reasoning
3044 behind this is that bare CRs in header lines are most likely either
3045 to be mistakes, or people trying to play silly games.
3047 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3048 was being incorrectly given as 18 bytes larger than it should have been.
3049 This is a VOB (very old bug).
3051 60. This may never have affected anything current, but just in case it has:
3052 When the local host is found other than at the start of a list of hosts,
3053 the local host, those with the same MX, and any that follow, are discarded.
3054 When the list in question was part of a longer list of hosts, the following
3055 hosts (not currently being processed) were also being discarded. This no
3056 longer happens. I'm not sure if this situation could ever has previously
3059 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3062 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3063 header. This is something that is recommended in a new Internet Draft, and
3064 is something that is documented as being done by Sendmail. There are two
3065 possible values. For messages generated by the autoreply transport, Exim
3068 Auto-Submitted: auto-replied
3070 whereas for all other generated messages (e.g. bounces) it adds
3072 Auto-Submitted: auto-generated
3074 63. The "personal" condition in filters now includes a test for the
3075 Auto-Submitted: header. If it contains the string "auto-" the message it
3076 not considered personal.
3078 64. Added rcpt_include_affixes as a generic transport option.
3080 65. Added queue_only_override (default true).
3082 66. Added the syslog_duplication option.
3084 67. If what should have been the first header line of a message consisted of
3085 a space followed by a colon, Exim was mis-interpreting it as a header line.
3086 It isn't of course - it is syntactically invalid and should therefore be
3087 treated as the start of the message body. The misbehaviour could have
3088 caused a number of strange effects, including loss of data in subsequent
3089 header lines, and spool format errors.
3091 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3092 client host had authenticated. This control can now be exercised by an ACL
3093 for more flexibility.
3095 69. By default, callouts do not happen when testing with -bh. There is now a
3096 variant, -bhc, which does actually run the callout code, including
3097 consulting and updating the callout cache.
3099 70. Added support for saslauthd authentication, courtesy of Alexander
3102 71. If statvfs() failed on the spool or log directories while checking their
3103 size for availability, Exim confusingly gave the error "space shortage".
3104 Furthermore, in debugging mode it crashed with a floating point exception.
3105 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3106 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3107 serious problem, Exim now writes to the main and panic logs when this
3108 happens, with details of the failure. It then refuses to accept the
3109 incoming message, giving the message "spool directory problem" or "log
3110 directory problem" with a 421 code for SMTP messages.
3112 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3113 0, 1, and 2 exist, because some OS complain for execs without them (see
3114 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3115 descriptors. However, the code omitted to check that the open succeeded,
3116 causing mysterious errors if for some reason the permissions on /dev/null
3117 got screwed. Now Exim writes a message to the main and panic logs, and
3118 bombs out if it can't open /dev/null.
3120 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3121 interact so that it is all more flexible. It is supposed to remain
3122 backwards compatible. Also added extra_local_interfaces.
3124 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3125 to bomb out with an assertion failure - to the client this appears as a
3126 connection drop. This problem occurs in the part of the code that was taken
3127 from the Samba project. Fortunately, the assertion is in a very simple
3128 function, so I have fixed this by reproducing the function inline in the
3129 one place where it is called, and arranging for authentication to fail
3130 instead of killing the process with assert().
3132 75. The SPA client code was not working when the server requested OEM rather
3133 than Unicode encoding.
3135 76. Added code to make require_files with a specific uid setting more usable in
3136 the case where statting the file as root fails - usually a non-root-mounted
3137 NFS file system. When this happens and the failure is EACCES, Exim now
3138 forks a subprocess and does the per-uid checking as the relevant uid.
3140 77. Added process_log_path.
3142 78. If log_file_path was not explicitly set, a setting of check_log_space or
3143 check_log_inodes was ignored.
3145 79. If a space check for the spool or log partitions fails, the incident is now
3146 logged. Of course, in the latter case the data may get lost...
3148 80. Added the %p formatting code to string_format() so that it can be used to
3149 print addresses in debug_print(). Adjusted all the address printing in the
3150 debugging in store.c to use %p rather than %d.
3152 81. There was a concern that a line of code in smtp_in.c could overflow a
3153 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3154 initially expressed, the concern was not well-founded, because trailing
3155 spaces are removed early. However, if the trailing spaces were followed by
3156 a NULL, they did not get removed, so the overflow was possible. Two fixes
3159 (a) I re-wrote the offending code in a cleaner fashion.
3160 (b) If an incoming SMTP command contains a NULL character, it is rejected
3163 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3164 runs initgroups(), so that if the Exim user is in any additional groups,
3165 they will be used during message reception.
3171 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob