1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.182 2005/07/23 20:59:16 tom 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.
16 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
18 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
19 command itself ended in error, the underlying message about the transport
20 filter timeout was being overwritten with the pipe command error. Now the
21 underlying error message should be appended to the second error message.
23 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
25 PH/02 Reduce the amount of output that "make" produces by default. Full output
26 can still be requested.
28 PH/03 The warning log line about a condition test deferring for a "warn" verb
29 was being output only once per connection, rather than after each
30 occurrence (because it was using the same function as for successful
31 "warn" verbs). This seems wrong, so I have changed it.
33 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
34 it should not have, which might have caused a crash in the right
35 circumstances, but probably never did.
37 PH/04 Installed a modified version of Tony Finch's patch to make submission
38 mode fix the return path as well as the Sender: header line, and to
39 add a /name= option so that you can make the user's friendly name appear
42 TF/03 Added the control = fakedefer ACL modifier.
44 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
45 Mark Lowes for thorough testing.
47 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
49 TK/03 Merged latest SRS patch from Miles Wilton.
51 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
52 with the definition in sysexits.h (which is #included earlier).
53 Fortunately, Exim does not actually use EX_OK. The code used to try to
54 preserve the sysexits.h value, by assumimg that macro definitions were
55 scanned for macro replacements. I have been disabused of this notion,
56 so now the code just undefines EX_OK before #including unistd.h.
58 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
59 in the smtp transport. When a block could not be written in a single
60 write() function, the timeout was being re-applied to each part-write.
61 This seems wrong - if the receiver was accepting one byte at a time it
62 would take for ever. The timeout is now adjusted when this happens. It
63 doesn't have to be particularly precise.
65 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
66 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
68 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
69 without a domain if the check to <postmaster@domain> fails.
71 SC/01 Eximstats: added -xls and the ability to specify output files
72 (patch written by Frank Heydlauf).
74 SC/02 Eximstats: use FileHandles for outputing results.
76 SC/03 Eximstats: allow any combination of xls, txt, and html output.
78 SC/04 Eximstats: fixed display of large numbers with -nvr option
80 SC/05 Eximstats: fixed merging of reports with empty tables.
82 SC/06 Eximstats: added the -include_original_destination flag
84 SC/07 Eximstats: removed tabs and trailing whitespace.
86 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
88 TK/06 MBOX spool code: Add real "From " MBOX separator line
89 so the .eml file is really in mbox format (even though
90 most programs do not really care). Patch from Alex Miller.
92 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
93 The latter is generated from $received_to and is only set if the
94 message has one envelope recipient. SA can use these headers,
95 obviously out-of-the-box. Patch from Alex Miller.
97 PH/08 The ${def test on a variable was returning false if the variable's
98 value was "0", contrary to what the specification has always said!
99 The result should be true unless the variable is empty.
101 PH/09 The syntax error of a character other than { following "${if
102 def:variable_name" (after optional whitespace) was not being diagnosed.
103 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
104 accidental colon was present, for example, could give incorrect results.
106 PH/10 Tidied the code in a number of places where the st_size field of a stat()
107 result is used (not including appendfile, where other changes are about
110 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
111 This involved changing a lot of size variables from int to off_t. It
112 should work with maildirs and everything.
114 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
115 spamd dying while we are connected to it.
117 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
120 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
121 function for BDB 4.3.
123 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
125 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
126 the sources for PCRE have been re-arranged and more files are now
129 PH/15 The code I had for printing potentially long long variables in PH/11
130 above was not the best (it lost precision). The length of off_t variables
131 is now inspected at build time, and an appropriate printing format (%ld
132 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
133 to be "long long int" or "long int". This is needed for the internal
134 formatting function string_vformat().
136 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
137 the configuration file to be ":syslog", then the script "guesses" where
138 the logs files are, rather than using the compiled in default. In our
139 case the guess is not the same as the compiled default, so the script
140 suddenly stopped working when I started to use syslog. The patch checks
141 to see if log_file_path is "". If so, it attempts to read it from exim
142 with no configuration file to get the compiled in version, before it
143 falls back to the previous guessing code."
145 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
146 implementing BATV in an Exim configuration. See NewStuff for the gory
149 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
150 Makefile that are specific to HP-UX.
152 PH/18 If the "use_postmaster" option was set for a recipient callout together
153 with the "random" option, the postmaster address was used as the MAIL
154 FROM address for the random test, but not for the subsequent recipient
155 test. It is now used for both.
157 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
158 patch removes a few documentation additions to RFC 3028, because the
159 latest draft now contains them. It adds the new en;ascii-case comparator
160 and a new error check for 8bit text in MIME parts. Comparator and
161 require names are now matched exactly. I enabled the subaddress
162 extension, but it is not well tested yet (read: it works for me)."
164 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
165 rework some of the code of TK/09 above to avoid the hardwired use of
166 "%lld" and "long long". Replaced the call to snprintf() with a call to
169 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
170 records point to non-existent hosts", "retry timeout exceeded", and
171 "retry time not reached for any host after a long failure period".
173 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
174 experimental DomainKeys support:
176 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
177 (2) On an error such as an illegally used "control", the wrong name for
178 the control was given.
180 These problems did NOT occur unless DomainKeys support was compiled.
182 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
184 PH/24 Added ${if match_ip condition.
186 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
187 databases so that it will be absolutely obvious if a crash occurs in the
188 DB library. This is a regular occurrence (often caused by mis-matched
191 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
192 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
193 picked up on a user's system that detects such things. There doesn't seem
194 to be a gcc warning option for this - only an attribute that has to be
195 put on the function's prototype. It seems that in Fedora Core 4 they have
196 set this on a number of new functions. No doubt there will be more in due
199 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
200 specify a transport. However, if an address that was verified by such a
201 router was the subject of a callout, Exim crashed because it tried to
202 read the rcpt_include_affixes from the non-existent transport. Now it
203 just assumes that the setting of that option is false. This bug was
204 introduced by 4.51/PH/31.
206 PH/28 Changed -d+all to exclude +memory, because that information is very
207 rarely of interest, but it makes the output a lot bigger. People tend to
208 do -d+all out of habit.
210 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
211 code in os-type was giving problems when libc.so lives in lib64, like on
214 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
215 aren't the modern standard, and it seems that some systems' include files
216 don't always have them. Exim was already checking for some of the newer
217 ones like T_AAAA, and defining it itself. I've added checks for all the
218 record types that Exim uses.
220 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
221 not automatically generating a new one, as it is supposed to. This
222 prevented TLS from working. If the file did exist, but contained invalid
223 data, a new version was generated, as expected. It was only the case of a
224 non-existent file that was broken.
226 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
227 with a change in libdomainkeys > 0.64.
229 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
230 from DNS. If the selector record carries the flag, it now has
231 precedence over the domain-wide flag.
233 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
235 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
236 the use of a port name that isn't defined in /etc/services) occurred, the
237 message was deferred as in a normal delivery, and thus remained on the
238 spool, instead of being failed because of the mua_wrapper setting. This
239 is now fixed, and I tidied up some of the mua_wrapper messages at the
242 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
243 the messages in a hash of arrays rather than using individual hashes.
244 This is a bit cleaner and results in dramatic memory savings, albeit
245 at a slight CPU cost.
247 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
248 as requested by Marc Sherman.
250 SC/10 Eximstats: added histograms for user specified patterns as requested
253 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
255 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
256 fopen() in the content-scanning modules that did not already have it.
262 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
263 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
265 TK/02 Fix ACL "control" statement not being available in MIME ACL.
267 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
269 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
270 to test Sieve filters that use "vacation".
272 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
273 that changes the way the GnuTLS parameters are stored in the cache file.
274 The new format can be generated externally. For backward compatibility,
275 if the data in the cache doesn't make sense, Exim assumes it has read an
276 old-format file, and it generates new data and writes a new file. This
277 means that you can't go back to an older release without removing the
280 PH/03 A redirect router that has both "unseen" and "one_time" set does not
281 work if there are any delivery delays because "one_time" forces the
282 parent to be marked "delivered", so its unseen clone is never tried
283 again. For this reason, Exim now forbids the simultaneous setting of
286 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
287 redirected to themselves ("homonym" addresses). Read the long ChangeLog
288 entry if you want to know the details. The fix, however, neglected to
289 consider the case when local delivery batching is involved. The test for
290 "previously delivered" was not happening when checking to see if an
291 address could be batched with a previous (undelivered) one; under
292 certain circumstances this could lead to multiple deliveries to the same
295 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
296 in its include files, and this causes problems building Exim.
298 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
299 header_syntax) but Exim was just ignoring anything given after a slash.
300 In particular, this caused confusion with an attempt to use "verify =
301 reverse_host_lookup/defer_ok". An error is now given when options are
302 supplied for verify items that do not have them. (Maybe reverse_host_
303 lookup should have a defer_ok option, but that's a different point.)
305 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
306 defined by RFC 821) to 2048, because there were problems with some AUTH
307 commands, and RFC 1869 says the size should be increased for extended
308 SMTP commands that take arguments.
310 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
313 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
314 "unknown" error; now it says that the functionality isn't in the binary.
316 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
317 an address' error message when a string expansion fails (syntax or
318 whatever). Otherwise the password may appear in the log. Following change
319 PH/42 below, there is no longer a chance of it appearing in a bounce
322 PH/11 Installed exipick version 20050225.0 from John Jetmore.
324 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
325 of its addresses was ever tried. (Bugzilla bug #2.)
327 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
328 the result incorrectly in the debug output. (It correctly added a newline
329 to what was transported.)
331 TF/01 Added $received_time.
333 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
334 commented out examples of how to interface to a virus scanner and to
335 SpamAssassin. Also added commented examples of av_scanner and
336 spamd_address settings.
338 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
339 and controls are allowed in which ACLs. There were a couple of minor
340 errors. Some of the entries in the conditions table (which is a table of
341 where they are NOT allowed) were getting very unwieldy; rewrote them as a
342 negation of where the condition IS allowed.
344 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
346 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
347 header file does not have a version number, so I've had to invent a new
348 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
349 API. The code is untested by me (my Linux distribution still has 0.3.2 of
350 radiusclient), but it was contributed by a Radius user.
352 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
353 files or directories, for OpenSSL.
355 PH/19 When an Exim process that is running as root has to create an Exim log
356 file, it does so in a subprocess that runs as exim:exim so as to get the
357 ownership right at creation (otherwise, other Exim processes might see
358 the file with the wrong ownership). There was no test for failure of this
359 fork() call, which would lead to the process getting stuck as it waited
360 for a non-existent subprocess. Forks do occasionally fail when resources
361 run out. I reviewed all the other calls to fork(); they all seem to check
364 PH/20 When checking for unexpected SMTP input at connect time (before writing
365 the banner), Exim was not dealing correctly with a non-positive return
366 from the read() function. If the client had disconnected by this time,
367 the result was a log entry for a synchronization error with an empty
368 string after "input=" when read() returned zero. If read() returned -1
369 (an event I could not check), uninitialized data bytes were printed.
370 There were reports of junk text (parts of files, etc) appearing after
373 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
375 PH/22 Added support for macro redefinition, and (re)definition in between
376 driver and ACL definitions.
378 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
379 forgetting to use the resulting value; it was using the unexpanded value.
381 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
382 hadn't been configured. The fix is from Juergen Kreileder, who
383 understands it better than I do:
385 "Here's what I see happening with three configured cyrus_sasl
386 authenticators configured (plain, login, cram-md5):
388 On startup auth_cyrus_sasl_init() gets called for each of these.
389 This means three calls to sasl_listmech() without a specified mech_list.
390 => SASL tests which mechs of all available mechs actually work
391 => three warnings about OTP not working
392 => the returned list contains: plain, login, cram-md5, digest-md5, ...
394 With the patch, sasl_listmech() also gets called three times. But now
395 SASL's mech_list option is set to the server_mech specified in the the
396 authenticator. Or in other words, the answer from sasl_listmech()
397 gets limited to just the mech you're testing for (which is different
399 => the return list contains just 'plain' or 'login', 'cram-md5' or
400 nothing depending on the value of ob->server_mech.
402 I've just tested the patch: Authentication still works fine,
403 unavailable mechs specified in the exim configuration are still
404 caught, and the auth.log warnings about OTP are gone."
406 PH/25 When debugging is enabled, the contents of the command line are added
407 to the debugging output, even when log_selector=+arguments is not
410 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
411 answer is "GNU", and only if the return is "GNU/something" is the answer
414 PH/27 $acl_verify_message is now set immediately after the failure of a
415 verification in an ACL, and so is available in subsequent modifiers. In
416 particular, the message can be preserved by coding like this:
418 warn !verify = sender
419 set acl_m0 = $acl_verify_message
421 Previously, $acl_verify_message was set only while expanding "message"
422 and "log_message" when a very denied access.
424 PH/28 Modified OS/os.c-Linux with
426 -#ifndef OS_LOAD_AVERAGE
427 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
429 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
430 nomenclature these days.)
432 PH/29 Installed patch from the Sieve maintainer that adds the options
433 sieve_useraddress and sieve_subaddress to the redirect router.
435 PH/30 In these circumstances:
436 . Two addresses routed to the same list of hosts;
437 . First host does not offer TLS;
438 . First host accepts first address;
439 . First host gives temporary error to second address;
440 . Second host offers TLS and a TLS session is established;
441 . Second host accepts second address.
442 Exim incorrectly logged both deliveries with the TLS parameters (cipher
443 and peerdn, if requested) that were in fact used only for the second
446 PH/31 When doing a callout as part of verifying an address, Exim was not paying
447 attention to any local part prefix or suffix that was matched by the
448 router that accepted the address. It now behaves in the same way as it
449 does for delivery: the affixes are removed from the local part unless
450 rcpt_include_affixes is set on the transport.
452 PH/32 Add the sender address, as F=<...>, to the log line when logging a
453 timeout during the DATA phase of an incoming message.
455 PH/33 Sieve envelope tests were broken for match types other than :is. I have
456 applied a patch sanctioned by the Sieve maintainer.
458 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
459 the uid or gid is negative. A case of a negative gid caused this to be
460 noticed. The fix allows for either to be negative.
462 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
463 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
466 PH/36 The change PH/12 above was broken. Fixed it.
468 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
469 the grounds that routing the same address twice would always produce the
470 same answer. This might have been true once, but it is certainly no
471 longer true now. Routing a child address may depend on the previous
472 routing that produced that child. Some complicated redirection strategies
473 went wrong when messages had multiple recipients, and made Exim's
474 behaviour dependent on the order in which the addresses were given.
476 I have moved the duplicate checking until after the routing is complete.
477 Exim scans the addresses that are assigned to local and remote
478 transports, and removes any duplicates. This means that more work will be
479 done, as duplicates will always all be routed, but duplicates are
480 presumably rare, so I don't expect this is of any significance.
482 For deliveries to pipes, files, and autoreplies, the duplicate checking
483 still happens during the routing process, since they are not going to be
486 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
487 It corrects a timeout issue with spamd. This is Ian's comment: "The
488 background is that sometimes spamd either never reads data from a
489 connection it has accepted, or it never writes response data. The exiscan
490 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
491 blindly assumes that writes won't block so it may never time out."
493 PH/39 Allow G after quota size as well as K and M.
495 PH/40 The value set for $authenticated_id in an authenticator may not contain
496 binary zeroes or newlines because the value is written to log lines and
497 to spool files. There was no check on this. Now the value is run through
498 the string_printing() function so that such characters are converted to
499 printable escape sequences.
501 PH/41 $message_linecount is a new variable that contains the total number of
502 lines in the message. Compare $body_linecount, which is the count for the
505 PH/42 Exim no longer gives details of delivery errors for specific addresses in
506 bounce and delay warning messages, except in certain special cases, which
509 (a) An SMTP error message from a remote host;
510 (b) A message specified in a :fail: redirection;
511 (c) A message specified in a "fail" command in a system filter;
512 (d) A message specified in a FAIL return from the queryprogram router;
513 (e) A message specified by the cannot_route_message router option.
515 In these cases only, Exim does include the error details in bounce and
516 warning messages. There are also a few cases where bland messages such
517 as "unrouteable address" or "local delivery error" are given.
519 PH/43 $value is now also set for the "else" part of a ${run expansion.
521 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
522 being worked on, but at least Exim now implements the latest version to
525 PH/45 In a pipe transport, although a timeout while waiting for the pipe
526 process to complete was treated as a delivery failure, a timeout while
527 writing the message to the pipe was logged, but erroneously treated as a
528 successful delivery. Such timeouts include transport filter timeouts. For
529 consistency with the overall process timeout, these timeouts are now
530 treated as errors, giving rise to delivery failures by default. However,
531 there is now a new Boolean option for the pipe transport called
532 timeout_defer, which, if set TRUE, converts the failures into defers for
533 both kinds of timeout. A transport filter timeout is now identified in
536 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
537 systems where "make" and "gmake" are different, calling "gmake" at top
538 level broke things. I've arranged for the value of $(MAKE) to be passed
539 from the Makefile to this script so that it can call the same version of
543 A note about Exim versions 4.44 and 4.50
544 ----------------------------------------
546 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
547 changes of various kinds. As a consequence, a big documentation update was
548 needed. This delayed the release for rather longer than seemed good, especially
549 in the light of a couple of (minor) security issues. Therefore, the changes
550 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
551 release. So 4.44 and 4.50 are in effect two different branches that both start
554 I have left the 4.50 change log unchanged; it contains all the changes since
555 4.43. The change log for 4.44 is below; many of its items are identical to
556 those for 4.50. This seems to be the most sensible way to preserve the
557 historical information.
563 1. Minor wording change to the doc/README.SIEVE file.
565 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
566 computation of the current number of files was incorrect.
568 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
569 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
572 4. Give more explanation in the error message when the command for a transport
573 filter fails to execute.
575 5. There are several places where Exim runs a non-Exim command in a
576 subprocess. The SIGUSR1 signal should be disabled for these processes. This
577 was being done only for the command run by the queryprogram router. It is
578 now done for all such subprocesses. The other cases are: ${run, transport
579 filters, and the commands run by the lmtp and pipe transports.
581 6. Added CONFIGURE_GROUP build-time option.
583 7. Some older OS have a limit of 256 on the maximum number of file
584 descriptors. Exim was using setrlimit() to set 1000 as a large value
585 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
586 systems. I've change it so that if it can't get 1000, it tries for 256.
588 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
589 was an oversight, and furthermore, ever since the addition of extra
590 controls (e.g. 4.43/32), the checks on when to allow different forms of
591 "control" were broken. There should now be diagnostics for all cases when a
592 control that does not make sense is encountered.
594 9. Added the /retain_sender option to "control=submission".
596 10. $recipients is now available in the predata ACL (oversight).
598 11. Tidy the search cache before the fork to do a delivery from a message
599 received from the command line. Otherwise the child will trigger a lookup
600 failure and thereby defer the delivery if it tries to use (for example) a
601 cached ldap connection that the parent has called unbind on.
603 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
604 of $address_data from the recipient verification was clobbered by the
607 13. The value of address_data from a sender verification is now available in
608 $sender_address_data in subsequent conditions in the ACL statement.
610 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
612 15. Added a new option "connect=<time>" to callout options, to set a different
615 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
616 was its contents. (It was OK if the option was not defined at all.)
618 17. A "Completed" log line is now written for messages that are removed from
619 the spool by the -Mrm option.
621 18. New variables $sender_verify_failure and $recipient_verify_failure contain
622 information about exactly what failed.
624 19. Added -dd to debug only the daemon process.
626 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
627 handles timeouts, both on the server side and network timeouts. Renamed the
628 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
630 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
631 It is now set to "smtps".
633 22. $host_address is now set to the target address during the checking of
636 23. When checking ignore_target_hosts for an ipliteral router, no host name was
637 being passed; this would have caused $sender_host_name to have been used if
638 matching the list had actually called for a host name (not very likely,
639 since this list is usually IP addresses). A host name is now passed as
642 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
643 code that specifies a non-restarting handler (typically sigaction() in
644 modern systems) in an attempt to fix a rare and obscure crash bug.
646 25. Narrowed the window for a race in the daemon that could cause it to ignore
647 SIGCHLD signals. This is not a major problem, because they are used only to
648 wake it up if nothing else does.
650 26. A malformed maildirsize file could cause Exim to calculate negative values
651 for the mailbox size or file count. Odd effects could occur as a result.
652 The maildirsize information is now recalculated if the size or filecount
655 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
656 support for a long time. Removed HAVE_SYS_VFS_H.
658 28. Installed the latest version of exipick from John Jetmore.
660 29. In an address list, if the pattern was not a regular expression, an empty
661 subject address (from a bounce message) matched only if the pattern was an
662 empty string. Non-empty patterns were not even tested. This was the wrong
663 because it is perfectly reasonable to use an empty address as part of a
664 database query. An empty address is now tested by patterns that are
665 lookups. However, all the other forms of pattern expect the subject to
666 contain a local part and a domain, and therefore, for them, an empty
667 address still always fails if the pattern is not itself empty.
669 30. Exim went into a mad DNS loop when attempting to do a callout where the
670 host was specified on an smtp transport, and looking it up yielded more
673 31. Re-factored the code for checking spool and log partition space into a
674 function that finds that data and another that does the check. The former
675 is then used to implement four new variables: $spool_space, $log_space,
676 $spool_inodes, and $log_inodes.
678 32. The RFC2047 encoding function was originally intended for short strings
679 such as real names; it was not keeping to the 75-character limit for
680 encoded words that the RFC imposes. It now respects the limit, and
681 generates multiple encoded words if necessary. To be on the safe side, I
682 have increased the buffer size for the ${rfc2047: expansion operator from
685 33. It is now permitted to omit both strings after an "if" condition; if the
686 condition is true, the result is "true". As before, when the second string
687 is omitted, a false condition yields an empty string. This makes it less
688 cumbersome to write custom ACL and router conditions.
690 34. Failure to deliver a bounce message always caused it to be frozen, even if
691 there was an errors_to setting on the router. The errors_to setting is now
694 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
695 canonical form (fully expanded) before being placed in
696 $sender_host_address.
698 36. The table in the code that translates DNS record types into text (T_A to
699 "A" for instance) was missing entries for NS and CNAME. It is just possible
700 that this could have caused confusion if both these types were looked up
701 for the same domain, because the text type is used as part of Exim's
702 per-process caching. But the chance of anyone hitting this buglet seems
705 37. The dnsdb lookup has been extended in a number of ways.
707 (1) There is a new type, "zns", which walks up the domain tree until it
708 finds some nameserver records. It should be used with care.
710 (2) There is a new type, "mxh", which is like "mx" except that it returns
711 just the host names, not the priorities.
713 (3) It is now possible to give a list of domains (or IP addresses) to be
714 looked up. The behaviour when one of the lookups defers can be
715 controlled by a keyword.
717 (4) It is now possible to specify the separator character for use when
718 multiple records are returned.
720 38. The dnslists ACL condition has been extended: it is now possible to supply
721 a list of IP addresses and/or domains to be looked up in a particular DNS
724 39. Added log_selector=+queue_time_overall.
726 40. When running the queue in the test harness, wait just a tad after forking a
727 delivery process, to get repeatability of debugging output.
729 41. Include certificate and key file names in error message when GnuTLS fails
730 to set them up, because the GnuTLS error message doesn't include the name
731 of the failing file when there is a problem reading it.
733 42. Allow both -bf and -bF in the same test run.
735 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
737 44. The "Exiscan patch" is now merged into the mainline Exim source.
739 45. Sometimes the final signoff response after QUIT could fail to get
740 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
741 < 0 before doing a fflush(). This bug looks as though it goes back to the
742 introduction of TLS in release 3.20, but "sometimes" must have been rare
743 because the tests only now provoked it.
745 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
746 (this can affect the format of dates).
748 47. exim_tidydb, when checking for the continued existence of a message for
749 which it has found a message-specific retry record, was not finding
750 messages that were in split spool directories. Consequently, it was
751 deleting retry records that should have stayed in existence.
753 48. Steve fixed some bugs in eximstats.
755 49. The SPA authentication driver was not abandoning authentication and moving
756 on to the next authenticator when an expansion was forced to fail,
757 contradicting the general specification for all authenticators. Instead it
758 was generating a temporary error. It now behaves as specified.
760 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
761 (the order specifies the preference for clients). The order is now AES256,
762 AES128, 3DES, ARCFOUR128.
764 51. Small patch to Sieve code - explicitly set From: when generating an
767 52. Exim crashed if a remote delivery caused a very long error message to be
768 recorded - for instance if somebody sent an entire SpamAssassin report back
769 as a large number of 550 error lines. This bug was coincidentally fixed by
770 increasing the size of one of Exim's internal buffers (big_buffer) that
771 happened as part of the Exiscan merge. However, to be on the safe side, I
772 have made the code more robust (and fixed the comments that describe what
775 53. Now that there can be additional text after "Completed" in log lines (if
776 the queue_time_overall log selector is set), a one-byte patch to exigrep
777 was needed to allow it to recognize "Completed" as not the last thing in
780 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
781 patch that reportedly fixes this has been added. I am not expert enough to
782 create a test for it. This is what the patch creator wrote:
784 "I found a little strange behaviour of ldap code when working with
785 Windows 2003 AD Domain, where users was placed in more than one
786 Organization Units. When I tried to give exim partial DN, the exit code
787 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
788 But simultaneously result of request was absolutely normal ldap result,
789 so I produce this patch..."
791 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
792 REFERENCE, so I have modified the code to exclude the patch when that macro
795 55. Some experimental protocols are using DNS PTR records for new purposes. The
796 keys for these records are domain names, not reversed IP addresses. The
797 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
798 leaves it alone. Component reversal etc. now happens only for IP addresses.
801 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
803 57. Double the size of the debug message buffer (to 2048) so that more of very
804 long debug lines gets shown.
806 58. The exicyclog utility now does better if the number of log files to keep
807 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
809 59. Two changes related to the smtp_active_hostname option:
811 (1) $smtp_active_hostname is now available as a variable.
812 (2) The default for smtp_banner uses $smtp_active_hostname instead
813 of $primary_hostname.
815 60. The host_aton() function is supposed to be passed a string that is known
816 to be a valid IP address. However, in the case of IPv6 addresses, it was
817 not checking this. This is a hostage to fortune. Exim now panics and dies
818 if the condition is not met. A case was found where this could be provoked
819 from a dnsdb PTR lookup with an IPv6 address that had more than 8
820 components; fortuitously, this particular loophole had already been fixed
821 by change 4.50/55 above.
823 If there are any other similar loopholes, the new check in host_aton()
824 itself should stop them being exploited. The report I received stated that
825 data on the command line could provoke the exploit when Exim was running as
826 exim, but did not say which command line option was involved. All I could
827 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
831 61. There was a buffer overflow vulnerability in the SPA authentication code
832 (which came originally from the Samba project). I have added a test to the
833 spa_base64_to_bits() function which I hope fixes it.
836 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
837 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
839 63. The daemon start-up calls getloadavg() while still root for those OS that
840 need the first call to be done as root, but it missed one case: when
841 deliver_queue_load_max is set with deliver_drop_privilege. This is
842 necessary for the benefit of the queue runner, because there is no re-exec
843 when deliver_drop_privilege is set.
845 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
848 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
849 was in use, was not putting the data itself into the right store pool;
850 consequently, it could be overwritten for a subsequent message in the same
851 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
854 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
856 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
857 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
858 the function were treating the return as a boolean value, which happened to
859 work because 0=false and not-0=true, but is not correct code.
861 68. The host_aton() function was not handling scoped IPv6 addresses (those
862 with, for example, "%eth0" on the end) correctly.
864 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
865 negated items (that is, ~something) in unsigned ints. Some compilers
866 apparently mutter when there is no cast.
868 70. If an address verification called from an ACL failed, and did not produce a
869 user-specific message (i.e. there was only a "system" message), nothing was
870 put in $acl_verify_message. In this situation, it now puts the system
873 71. Change 4.23/11 added synchronization checking at the start of an SMTP
874 session; change 4.31/43 added the unwanted input to the log line - except
875 that it did not do this in the start of session case. It now does.
877 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
878 This is wrong and can cause the other end to generate a synchronization
879 error if it is another Exim or anything else that does the synchronization
880 check. A QUIT command is no longer sent after a timeout.
882 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
885 74. The defer_ok option of callout verification was not working if it was used
886 when verifying addresses in header lines, that is, for this case:
888 verify = header_sender/callout=defer_ok
890 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
891 those file descriptors could be used for SMTP connections. If anything
892 wrote to stderr (the example that came up was "warn" in embedded Perl), it
893 could be sent to the SMTP client, causing chaos. The daemon now opens
894 stdin, stdout, and stderr to /dev/null when it puts itself into the
897 76. Arrange for output from Perl's "warn" command to be written to Exim's main
898 log by default. The user can override this with suitable Perl magic.
900 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
901 the log message when discard triggers, was not working for the DATA ACL or
902 for the non-SMTP ACL.
904 78. Error message wording change in sieve.c.
906 79. If smtp_accept_max_per_host was set, the number of connections could be
907 restricted to fewer than expected, because the daemon was trying to set up
908 a new connection before checking whether the processes handling previous
909 connections had finished. The check for completed processes is now done
910 earlier. On busy systems, this bug wouldn't be noticed because something
911 else would have woken the daemon, and it would have reaped the completed
914 80. If a message was submitted locally by a user whose login name contained one
915 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
916 It caused a spool format error. I have fixed the spool reading code. A
917 related problem was that the "from" clause in the Received: line became
918 illegal because of the space(s). It is now covered by ${quote_local_part.
920 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
923 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
925 83. Patch from Sieve maintainer to fix unterminated string problem in
928 84. Some minor changes to the Linux configuration files to help with other
929 OS variants using glibc.
931 85. One more patch for Sieve to update vacation handling to latest spec.
934 ----------------------------------------------------
935 See the note above about the 4.44 and 4.50 releases.
936 ----------------------------------------------------
942 1. Change 4.43/35 introduced a bug that caused file counts to be
943 incorrectly computed when quota_filecount was set in an appendfile
946 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
947 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
950 3. Give more explanation in the error message when the command for a transport
951 filter fails to execute.
953 4. There are several places where Exim runs a non-Exim command in a
954 subprocess. The SIGUSR1 signal should be disabled for these processes. This
955 was being done only for the command run by the queryprogram router. It is
956 now done for all such subprocesses. The other cases are: ${run, transport
957 filters, and the commands run by the lmtp and pipe transports.
959 5. Some older OS have a limit of 256 on the maximum number of file
960 descriptors. Exim was using setrlimit() to set 1000 as a large value
961 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
962 systems. I've change it so that if it can't get 1000, it tries for 256.
964 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
965 was an oversight, and furthermore, ever since the addition of extra
966 controls (e.g. 4.43/32), the checks on when to allow different forms of
967 "control" were broken. There should now be diagnostics for all cases when a
968 control that does not make sense is encountered.
970 7. $recipients is now available in the predata ACL (oversight).
972 8. Tidy the search cache before the fork to do a delivery from a message
973 received from the command line. Otherwise the child will trigger a lookup
974 failure and thereby defer the delivery if it tries to use (for example) a
975 cached ldap connection that the parent has called unbind on.
977 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
978 of $address_data from the recipient verification was clobbered by the
981 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
982 was its contents. (It was OK if the option was not defined at all.)
984 11. A "Completed" log line is now written for messages that are removed from
985 the spool by the -Mrm option.
987 12. $host_address is now set to the target address during the checking of
990 13. When checking ignore_target_hosts for an ipliteral router, no host name was
991 being passed; this would have caused $sender_host_name to have been used if
992 matching the list had actually called for a host name (not very likely,
993 since this list is usually IP addresses). A host name is now passed as
996 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
997 code that specifies a non-restarting handler (typically sigaction() in
998 modern systems) in an attempt to fix a rare and obscure crash bug.
1000 15. Narrowed the window for a race in the daemon that could cause it to ignore
1001 SIGCHLD signals. This is not a major problem, because they are used only to
1002 wake it up if nothing else does.
1004 16. A malformed maildirsize file could cause Exim to calculate negative values
1005 for the mailbox size or file count. Odd effects could occur as a result.
1006 The maildirsize information is now recalculated if the size or filecount
1009 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1010 support for a long time. Removed HAVE_SYS_VFS_H.
1012 18. Updated exipick to current release from John Jetmore.
1014 19. Allow an empty sender to be matched against a lookup in an address list.
1015 Previously the only cases considered were a regular expression, or an
1018 20. Exim went into a mad DNS lookup loop when doing a callout where the
1019 host was specified on the transport, if the DNS lookup yielded more than
1022 21. The RFC2047 encoding function was originally intended for short strings
1023 such as real names; it was not keeping to the 75-character limit for
1024 encoded words that the RFC imposes. It now respects the limit, and
1025 generates multiple encoded words if necessary. To be on the safe side, I
1026 have increased the buffer size for the ${rfc2047: expansion operator from
1029 22. Failure to deliver a bounce message always caused it to be frozen, even if
1030 there was an errors_to setting on the router. The errors_to setting is now
1033 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1034 canonical form (fully expanded) before being placed in
1035 $sender_host_address.
1037 24. Updated eximstats to version 1.33
1039 25. Include certificate and key file names in error message when GnuTLS fails
1040 to set them up, because the GnuTLS error message doesn't include the name
1041 of the failing file when there is a problem reading it.
1043 26. Expand error message when OpenSSL has problems setting up cert/key files.
1046 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1047 (this can affect the format of dates).
1049 28. exim_tidydb, when checking for the continued existence of a message for
1050 which it has found a message-specific retry record, was not finding
1051 messages that were in split spool directories. Consequently, it was
1052 deleting retry records that should have stayed in existence.
1054 29. eximstats updated to version 1.35
1055 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1056 1.35 - bugfix such that pie charts by volume are generated correctly
1058 30. The SPA authentication driver was not abandoning authentication and moving
1059 on to the next authenticator when an expansion was forced to fail,
1060 contradicting the general specification for all authenticators. Instead it
1061 was generating a temporary error. It now behaves as specified.
1063 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1064 (the order specifies the preference for clients). The order is now AES256,
1065 AES128, 3DES, ARCFOUR128.
1067 31. Small patch to Sieve code - explicitly set From: when generating an
1070 32. Exim crashed if a remote delivery caused a very long error message to be
1071 recorded - for instance if somebody sent an entire SpamAssassin report back
1072 as a large number of 550 error lines. This bug was coincidentally fixed by
1073 increasing the size of one of Exim's internal buffers (big_buffer) that
1074 happened as part of the Exiscan merge. However, to be on the safe side, I
1075 have made the code more robust (and fixed the comments that describe what
1078 33. Some experimental protocols are using DNS PTR records for new purposes. The
1079 keys for these records are domain names, not reversed IP addresses. The
1080 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1081 leaves it alone. Component reversal etc. now happens only for IP addresses.
1084 34. The host_aton() function is supposed to be passed a string that is known
1085 to be a valid IP address. However, in the case of IPv6 addresses, it was
1086 not checking this. This is a hostage to fortune. Exim now panics and dies
1087 if the condition is not met. A case was found where this could be provoked
1088 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1089 components; fortuitously, this particular loophole had already been fixed
1090 by change 4.50/55 or 4.44/33 above.
1092 If there are any other similar loopholes, the new check in host_aton()
1093 itself should stop them being exploited. The report I received stated that
1094 data on the command line could provoke the exploit when Exim was running as
1095 exim, but did not say which command line option was involved. All I could
1096 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1097 running as the user.
1100 35. There was a buffer overflow vulnerability in the SPA authentication code
1101 (which came originally from the Samba project). I have added a test to the
1102 spa_base64_to_bits() function which I hope fixes it.
1105 36. The daemon start-up calls getloadavg() while still root for those OS that
1106 need the first call to be done as root, but it missed one case: when
1107 deliver_queue_load_max is set with deliver_drop_privilege. This is
1108 necessary for the benefit of the queue runner, because there is no re-exec
1109 when deliver_drop_privilege is set.
1111 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1112 was in use, was not putting the data itself into the right store pool;
1113 consequently, it could be overwritten for a subsequent message in the same
1114 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1117 38. Sometimes the final signoff response after QUIT could fail to get
1118 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1119 < 0 before doing a fflush(). This bug looks as though it goes back to the
1120 introduction of TLS in release 3.20, but "sometimes" must have been rare
1121 because the tests only now provoked it.
1127 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1128 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1129 Now it returns an integer. A number of calls were still expecting a T/F
1130 return. Fortuitously, in all cases, the tests worked in OK situations,
1131 which is the norm. However, things would have gone wrong on any write
1132 failures on the smtp file descriptor. This function is used when sending
1133 messages over SMTP and also when doing verify callouts.
1135 2. When Exim is called to do synchronous delivery of a locally submitted
1136 message (the -odf or -odi options), it no longer closes stderr before doing
1139 3. Implemented the mua_wrapper option.
1141 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1143 5. Implemented the functions header_remove(), header_testname(),
1144 header_add_at_position(), and receive_remove_recipient(), and exported them
1147 6. If an ACL "warn" statement specified the addition of headers, Exim already
1148 inserted X-ACL-Warn: at the start if there was no header name. However, it
1149 was not making this test for the second and subsequent header lines if
1150 there were newlines in the string. This meant that an invalid header could
1151 be inserted if Exim was badly configured.
1153 7. Allow an ACL "warn" statement to add header lines at the start or after all
1154 the Received: headers, as well as at the end.
1156 8. Added the rcpt_4xx retry error code.
1158 9. Added postmaster_mailfrom=xxx to callout verification option.
1160 10. Added mailfrom=xxxx to the callout verification option, for verify=
1163 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1164 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1166 12. Inserted some casts to stop certain compilers warning when using pointer
1167 differences as field lengths or precisions in printf-type calls (mostly
1168 affecting debugging statements).
1170 13. Added optional readline() support for -be (dynamically loaded).
1172 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1173 same clock tick as a message's arrival, so that its received time was the
1174 same as the "first fail" time on the retry record, and that message
1175 remained on the queue past the ultimate address timeout, every queue runner
1176 would try a delivery (because it was past the ultimate address timeout) but
1177 after another failure, the ultimate address timeout, which should have then
1178 bounced the address, did not kick in. This was a "< instead of <=" error;
1179 in most cases the first failure would have been in the next clock tick
1180 after the received time, and all would be well.
1182 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1183 being recognized when the domain list was tested by the match_domain
1184 condition in an expansion string.
1186 16. Added the ${str2b64: operator.
1188 17. Exim was always calling setrlimit() to set a large limit for the number of
1189 processes, without checking whether the existing limit was already
1190 adequate. (It did check for the limit on file descriptors.) Furthermore,
1191 errors from getrlimit() and setrlimit() were being ignored. Now they are
1192 logged to the main and panic logs, but Exim does carry on, to try to do its
1193 job under whatever limits there are.
1195 18. Imported PCRE 5.0.
1197 19. Trivial typo in log message " temporarily refused connection" (the leading
1200 20. If the log selector return_path_on_delivery was set and an address was
1201 redirected to /dev/null, the delivery process crashed because it assumed
1202 that a return path would always be set for a "successful" delivery. In this
1203 case, the whole delivery is bypassed as an optimization, and therefore no
1206 21. Internal re-arrangement: the function for sending a challenge and reading
1207 a response while authentication was assuming a zero-terminated challenge
1208 string. It's now changed to take a pointer and a length, to allow for
1209 binary data in such strings.
1211 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1213 23. Exim was not respecting finduser_retries when seeking the login of the
1214 uid under which it was called; it was always trying 10 times. (The default
1215 setting of finduser_retries is zero.) Also, it was sleeping after the final
1216 failure, which is pointless.
1218 24. Implemented tls_on_connect_ports.
1220 25. Implemented acl_smtp_predata.
1222 26. If the domain in control=submission is set empty, Exim assumes that the
1223 authenticated id is a complete email address when it generates From: or
1224 Sender: header lines.
1226 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1227 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1228 chown and chgrp in /bin and hostname in /usr/bin.
1230 28. Exim was keeping the "process log" file open after each use, just as it
1231 does for the main log. This opens the possibility of it remaining open for
1232 long periods when the USR1 signal hits a daemon. Occasional processlog
1233 errors were reported, that could have been caused by this. Anyway, it seems
1234 much more sensible not to leave this file open at all, so that is what now
1237 29. The long-running daemon process does not normally write to the log once it
1238 has entered its main loop, and it closes the log before doing so. This is
1239 so that log files can straightforwardly be renamed and moved. However,
1240 there are a couple of unusual error situations where the daemon does write
1241 log entries, and I had neglected to close the log afterwards.
1243 30. The text of an SMTP error response that was received during a remote
1244 delivery was being truncated at 512 bytes. This is too short for some of
1245 the long messages that one sometimes sees. I've increased the limit to
1248 31. It is now possible to make retry rules that apply only when a message has a
1249 specific sender, in particular, an empty sender.
1251 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1252 it possible to be selective about when SMTP synchronization is enforced.
1254 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1256 32. Implemented hosts_connection_nolog.
1258 33. Added an ACL for QUIT.
1260 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1263 35. Added mailbox_size and mailbox_filecount to appendfile.
1265 36. Added control = no_multiline_responses to ACLs.
1267 37. There was a bug in the logic of the code that waits for the clock to tick
1268 in the case where the clock went backwards by a substantial amount such
1269 that the microsecond fraction of "now" was more than the microsecond
1270 fraction of "then" (but the whole seconds number was less).
1272 38. Added support for the libradius Radius client library this is found on
1273 FreeBSD (previously only the radiusclient library was supported).
1279 1. When certain lookups returned multiple values in the form name=value, the
1280 quoting of the values was not always being done properly. Specifically:
1281 (a) If the value started with a double quote, but contained no whitespace,
1283 (b) If the value contained whitespace other than a space character (i.e.
1284 tabs or newlines or carriage returns) it was not quoted.
1285 This fix has been applied to the mysql and pgsql lookups by writing a
1286 separate quoting function and calling it from the lookup code. The fix
1287 should probably also be applied to nisplus, ibase and oracle lookups, but
1288 since I cannot test any of those, I have not disturbed their existing code.
1290 2. A hit in the callout cache for a specific address caused a log line with no
1291 reason for rejecting RCPT. Now it says "Previous (cached) callout
1292 verification failure".
1294 3. There was an off-by-one bug in the queryprogram router. An over-long
1295 return line was truncated at 256 instead of 255 characters, thereby
1296 overflowing its buffer with the terminating zero. As well as fixing this, I
1297 have increased the buffer size to 1024 (and made a note to document this).
1299 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1300 when Exim is waiting for an SMTP response from a remote server, Exim
1301 restarts its select() call on the socket, thereby resetting its timeout.
1302 This is not a problem when such interrupts are rare. Somebody set up a cron
1303 job to run exiwhat every 2 minutes, which is less than the normal select()
1304 timeout (5 or 10 minutes). This meant that the select() timeout never
1305 kicked in because it was always reset. I have fixed this by comparing the
1306 time when an interrupt arrives with the time at the start of the first call
1307 to select(). If more time than the timeout has elapsed, the interrupt is
1308 treated as a timeout.
1310 5. Some internal re-factoring in preparation for the addition of Sieve
1311 extensions (by MH). In particular, the "personal" test is moved to a
1312 separate function, and given an option for scanning Cc: and Bcc: (which is
1313 not set for Exim filters).
1315 6. When Exim created an email address using the login of the caller as the
1316 local part (e.g. when creating a From: or Sender: header line), it was not
1317 quoting the local part when it contained special characters such as @.
1319 7. Installed new OpenBSD configuration files.
1321 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1322 try to make them clearer.
1324 9. Callout options, other than the timeout value, were being ignored when
1325 verifying sender addresses in header lines. For example, when using
1327 verify = header_sender/callout=no_cache
1329 the cache was (incorrectly) being used.
1331 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1332 only the Cygwin environment.
1334 11. When return_path_on_delivery was set as a log selector, if different remote
1335 addresses in the same message used different return paths and parallel
1336 remote delivery occurred, the wrong values would sometimes be logged.
1337 (Whenever a remote delivery process finished, the return path value from
1338 the most recently started remote delivery process was logged.)
1340 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1341 lines when AUTH and/or TLS are in use. This is the "received protocol"
1342 field. Exim used to use "asmtp" for authenticated SMTP, without any
1343 indication (in the protocol name) for TLS use. Now it follows the RFC and
1344 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1345 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1346 names appear in log lines as well as in Received: header lines.
1348 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1349 extensions, and comparison tests, and to fix some bugs.
1351 14. Changes to the "personal" filter test:
1353 (1) The test was buggy in that it was just doing the equivalent of
1354 "contains" tests on header lines. For example, if a user's address was
1355 anne@some.where, the "personal" test would incorrectly be true for
1357 To: susanne@some.where
1359 This test is now done by extracting each address from the header in turn,
1360 and checking the entire address. Other tests that are part of "personal"
1361 are now done using regular expressions (for example, to check local parts
1362 of addresses in From: header lines).
1364 (2) The list of non-personal local parts in From: addresses has been
1365 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1366 taken from the Sieve specification recommendations.
1368 (3) If the message contains any header line starting with "List-" it is
1369 treated as non-personal.
1371 (4) The test for "circular" in the Subject: header line has been removed
1372 because it now seems ill-conceived.
1374 15. Minor typos in src/EDITME comments corrected.
1376 16. Installed latest exipick from John Jetmore.
1378 17. If headers_add on a router specified a text string that was too long for
1379 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1380 of string_sprintf() is now avoided.
1382 18. $message_body_size was not set (it was always zero) when running the DATA
1383 ACL and the local_scan() function.
1385 19. For the "mail" command in an Exim filter, no default was being set for
1386 the once_repeat time, causing a random time value to be used if "once" was
1387 specified. (If the value happened to be <= 0, no repeat happened.) The
1388 default is now 0s, meaning "never repeat". The "vacation" command was OK
1389 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1390 (I found it when inspecting the code).
1392 20. There is now an overall timeout for performing a callout verification. It
1393 defaults to 4 times the callout timeout, which applies to individual SMTP
1394 commands during the callout. The overall timeout applies when there is more
1395 than one host that can be tried. The timeout is checked before trying the
1396 next host. This prevents very long delays if there are a large number of
1397 hosts and all are timing out (e.g. when the network connections are timing
1398 out). The value of the overall timeout can be changed by specifying an
1399 additional sub-option for "callout", called "maxwait". For example:
1401 verify = sender/callout=5s,maxwait=20s
1403 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1404 to the end before writing, but this should make it even safer).
1406 22. Exim was forgetting that it had advertised PIPELINING for the second and
1407 subsequent messages on an SMTP connection. It was also not resetting its
1408 memory on STARTTLS and an internal HELO.
1410 23. When Exim logs an SMTP synchronization error within a session, it now
1411 records whether PIPELINING has been advertised or not.
1413 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1414 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1415 rather than long int.
1417 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1419 26. Added the never_mail option to autoreply.
1425 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1426 crash if the getsockname() call failed; this can happen if a connection is
1427 closed very soon after it is established. The problem was simply in the
1428 order in which certain operations were done, causing Exim to try to write
1429 to the SMTP stream before it had set up the file descriptor. The bug has
1430 been fixed by making things happen in the correct order.
1436 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1437 before the connection was closed, thus losing the rejection response.
1439 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1440 some early Solaris releases, but causes trouble in current releases where
1441 socklen_t is defined.
1443 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1446 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1449 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1450 that releasing the top of it at the end releases what was used for sub-
1451 expansions (unless the block got too big). However, discard this block if
1452 the first thing is a variable or header, so that we can use its block when
1453 it is dynamic (useful for very large $message_headers, for example).
1455 6. Lookups now cache *every* query, not just the most recent. A new, separate
1456 store pool is used for this. It can be recovered when all lookup caches are
1457 flushed. Lookups now release memory at the end of their result strings.
1458 This has involved some general refactoring of the lookup sources.
1460 7. Some code has been added to the store_xxx() functions to reduce the amount
1461 of flapping under certain conditions.
1463 8. log_incoming_interface used to affect only the <= reception log lines. Now
1464 it causes the local interface and port to be added to several more SMTP log
1465 lines, for example "SMTP connection from", and rejection lines.
1467 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1469 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1471 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1472 could be overwritten at the end of the current message (or the start of a
1473 new message if it was set in a HELO ACL). The value is now preserved for
1474 the duration of the SMTP connection.
1476 12. If a transport had a headers_rewrite setting, and a matching header line
1477 contained an unqualified address, that address was qualified, even if it
1478 did not match any rewriting rules. The underlying bug was that the values
1479 of the flags that permit the existence of unqualified sender and recipient
1480 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1481 non-local messages, and by -bnq for local messages) were not being
1482 preserved with the message after it was received.
1484 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1485 "next input=" as part of the text comprising the host identity instead of
1486 the correct text. The code was using the same buffer for two different
1487 strings. However, depending on which order the printing function evaluated
1488 its arguments, the bug did not always show up. Under Linux, for example, my
1489 test suite worked just fine.
1491 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1492 doesn't work with some older versions of Perl. It has been changed to "my",
1493 which in any case is probably the better facility to use.
1495 15. A really picky compiler found some instances of statements for creating
1496 error messages that either had too many or two few arguments for the format
1499 16. The size of the buffer for calls to the DNS resolver has been increased
1500 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1501 for addresses that have a lot of PTR records. This alleviates a problem; it
1502 does not fully solve it.
1504 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1505 buffer now truncates the list and logs the incident, which is the same
1506 action as happens when Exim is looking up a host name and its aliases.
1507 Previously in this situation something unpredictable would happen;
1508 sometimes it was "internal error: store_reset failed".
1510 18. If a server dropped the connection unexpectedly when an Exim client was
1511 using GnuTLS and trying to read a response, the client delivery process
1512 crashed while trying to generate an error log message.
1514 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1515 string, for example:
1517 warn message = H1: something\nH2: something
1519 the text was added as a single header line from Exim's point of view
1520 though it ended up OK in the delivered message. However, searching for the
1521 second and subsequent header lines using $h_h2: did not work. This has been
1522 fixed. Similarly, if a system filter added multiple headers in this way,
1523 the routers could not see them.
1525 20. Expanded the error message when iplsearch is called with an invalid key to
1526 suggest using net-iplsearch in a host list.
1528 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1529 ACLs are no longer actually imposed (and a message to that effect is
1532 22. If a "gecos" field in a passwd entry contained escaped characters, in
1533 particular, if it contained a \" sequence, Exim got it wrong when building
1534 a From: or a Sender: header from that name. A second bug also caused
1535 incorrect handling when an unquoted " was present following a character
1536 that needed quoting.
1538 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1539 was not being matched caselessly.
1541 24. Arranged for all hyphens in the exim.8 source to be escaped with
1544 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1545 properly. Recipient callout cache records were still being keyed to include
1546 the sender, even when use_sender was set false. This led to far more
1547 callouts that were necessary. The sender is no longer included in the key
1548 when use_sender is false.
1550 26. Added "control = submission" modifier to ACLs.
1552 27. Added the ${base62d: operator to decode base 62 numbers.
1554 28. dnsdb lookups can now access SRV records.
1556 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1557 the configuration file.
1559 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1560 (-v) mode. This makes the output for a verbose queue run more intelligible.
1562 31. Added a use_postmaster feature to recipient callouts.
1564 32. Added the $body_zerocount variable, containing the number of binary zero
1565 bytes in the message body.
1567 33. The time of last modification of the "new" subdirectory is now used as the
1568 "mailbox time last read" when there is a quota error for a maildir
1571 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1573 35. Added +ignore_unknown as a special item in host lists.
1575 36. Code for decoding IPv6 addresses in host lists is now included, even if
1576 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1577 address was recognized as an IP address, but was then not correctly decoded
1578 into binary, causing unexpected and incorrect effects when compared with
1585 1. Very minor rewording of debugging text in manualroute to say "list of
1586 hosts" instead of "hostlist".
1588 2. If verify=header_syntax was set, and a header line with an unqualified
1589 address (no domain) and a large number of spaces between the end of the
1590 name and the colon was received, the reception process suffered a buffer
1591 overflow, and (when I tested it) crashed. This was caused by some obsolete
1592 code that should have been removed. The fix is to remove it!
1594 3. When running in the test harness, delay a bit after writing a bounce
1595 message to get a bit more predictability in the log output.
1597 4. Added a call to search_tidyup() just before forking a reception process. In
1598 theory, someone could use a lookup in the expansion of smtp_accept_max_
1599 per_host which, without the tidyup, could leave open a database connection.
1601 5. Added the variables $recipient_data and $sender_data which get set from a
1602 lookup success in an ACL "recipients" or "senders" condition, or a router
1603 "senders" option, similar to $domain_data and $local_part_data.
1605 6. Moved the writing of debug_print from before to after the "senders" test
1608 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1609 problems for message scanning, either using a data ACL, or using
1610 local_scan() because the Received: header was not generated till after they
1611 were called (in order to set the time as the time of reception completion).
1612 I have revised the way this works. The header is now generated after the
1613 body is received, but before the ACL or local_scan() are called. After they
1614 are run, the timestamp in the header is updated.
1620 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1621 before starting a queue runner without re-exec. This happened only when
1622 deliver_drop_privilege was set or when the Exim user was set to root. The
1623 effect of the bug was that timeouts during subsequent deliveries caused
1624 crashes instead of being properly handled. The handler is now left at its
1625 default (and expected) setting.
1627 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1628 message, again when deliver_drop_privilege is set or Exim is run as root.
1629 The bug described in (1) was not present in this case, but the tidying up
1630 of the other signals was missing. I have made the two cases consistent.
1632 3. The ignore_target_hosts setting on a manualroute router was being ignored
1633 for hosts that were looked up using the /MX notation.
1635 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1638 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1639 operated on the sender address. After changing the $sender_address to <>
1640 for the sender address verify, Exim was re-instated it as the original
1641 (before rewriting) address, but remembering that it had rewritten it, so it
1642 wasn't rewriting it again. This bug also had the effect of breaking the
1643 sender address verification caching when the sender address was rewritten.
1645 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1646 This has been changed so that if the ip literal address matches
1647 ignore_target_hosts, the router declines.
1649 7. Added expansion conditions match_domain, match_address, and match_local_
1650 part (NOT match_host).
1652 8. The placeholder for the Received: header didn't have a length field set.
1654 9. Added code to Exim itself and to exim_lock to test for a specific race
1655 condition that could lead to file corruption when using MBX delivery. The
1656 issue is with the lockfile that is created in /tmp. If this file is removed
1657 after a process has opened it but before that process has acquired a lock,
1658 there is the potential for a second process to recreate the file and also
1659 acquire a lock. This could lead to two Exim processes writing to the file
1660 at the same time. The added code performs the same test as UW imapd; it
1661 checks after acquiring the lock that its file descriptor still refers to
1662 the same named file.
1664 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1665 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1668 11. Added the smtp_active_hostname option. If used, this will typically be made
1669 to depend on the incoming interface address. Because $interface_address is
1670 not set up until the daemon has forked a reception process, error responses
1671 that can happen earlier (such as "too many connections") no longer contain
1674 12. If an expansion in a condition on a "warn" statement fails because a lookup
1675 defers, the "warn" statement is abandoned, and the next ACL statement is
1676 processed. Previously this caused the whole ACL to be aborted.
1678 13. Added the iplsearch lookup type.
1680 14. Added ident_timeout as a log selector.
1682 15. Added tls_certificate_verified as a log selector.
1684 16. Added a global option tls_require_ciphers (compare the smtp transport
1685 option of the same name). This controls incoming TLS connections.
1687 17. I finally figured out how to make tls_require_ciphers do a similar thing
1688 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1689 before starting the TLS session.
1691 18. Tabs are now shown as \t in -bP output.
1693 19. If the log selector return_path_on_delivery was set, Exim crashed when
1694 bouncing a message because it had too many Received: header lines.
1696 20. If two routers both had headers_remove settings, and the first one included
1697 a superfluous trailing colon, the final name in the first list and the
1698 first name in the second list were incorrectly joined into one item (with a
1699 colon in the middle).
1705 1. Added -C and -D options to the exinext utility, mainly to make it easier
1706 to include in the automated testing, but these could be helpful when
1707 multiple configurations are in use.
1709 2. The exinext utility was not formatting the output nicely when there was
1710 an alternate port involved in the retry record key, nor when there was a
1711 message id as well (for retries that were specific to a specific message
1712 and a specific host). It was also confused by IPv6 addresses, because of
1713 the additional colons they contain. I have fixed the IPv4 problem, and
1714 patched it up to do a reasonable job for IPv6.
1716 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1717 delivery, the log line now contains "pipelined" if PIPELINING was used.
1719 4. An SMTP transport process used to panic and die if the bind() call to set
1720 an explicit outgoing interface failed. This has been changed; it is now
1721 treated in the same way as a connect() failure.
1723 5. A reference to $sender_host_name in the part of a conditional expansion
1724 that was being skipped was still causing a DNS lookup. This no longer
1727 6. The def: expansion condition was not recognizing references to header lines
1728 that used bh_ and bheader_.
1730 7. Added the _cache feature to named lists.
1732 8. The code for checking quota_filecount in the appendfile transport was
1733 allowing one more file than it should have been.
1735 9. For compatibility with Sendmail, the command line option
1743 and sets the incoming protocol and host name (for trusted callers). The
1744 host name and its colon can be omitted when only the protocol is to be set.
1745 Note the Exim already has two private options, -pd and -ps, that refer to
1746 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1747 "s", but I don't think that's a major issue.
1749 10. A number of refactoring changes to the code, none of which should affect
1752 (a) The number of logging options was getting close to filling up the
1753 32-bit word that was used as a bit map. I have split them into two classes:
1754 those that are passed in the argument to log_write(), and those that are
1755 only ever tested independently outside of that function. These are now in
1756 separate 32-bit words, so there is plenty of room for expansion again.
1757 There is no change in the user interface or the logging behaviour.
1759 (b) When building, for example, log lines, the code previously used a
1760 macro that called string_cat() twice, in order to add two strings. This is
1761 not really sufficiently general. Furthermore, there was one instance where
1762 it was actually wrong because one of the argument was used twice, and in
1763 one call a function was used. (As it happened, calling the function twice
1764 did not affect the overall behaviour.) The macro has been replaced by a
1765 function that can join an arbitrary number of extra strings onto a growing
1768 (c) The code for expansion conditions now uses a table and a binary chop
1769 instead of a serial search (which was left over from when there were very
1770 few conditions). Also, it now recognizes conditions like "pam" even when
1771 the relevant support is not compiled in: a suitably worded error message is
1772 given if an attempt is made to use such a condition.
1774 11. Added ${time_interval:xxxxx}.
1776 12. A bug was causing one of the ddress fields not to be passed back correctly
1777 from remote delivery subprocesses. The field in question was not being
1778 subsequently used, so this caused to problems in practice.
1780 13. Added new log selectors queue_time and deliver_time.
1782 14. Might have fixed a bug in maildirsizefile handling that threw up
1783 "unexpected character" debug warnings, and recalculated the data
1784 unnecessarily. In any case, I expanded the warning message to give more
1787 15. Added the message "Restricted characters in address" to the statements in
1788 the default ACL that block characters like @ and % in local parts.
1790 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1791 Three changes have been made:
1793 (a) There was a serious bug; a negative response to MAIL caused the whole
1794 recipient domain to be cached as invalid, thereby blocking all messages
1795 to all local parts at the same domain, from all senders. This bug has
1796 been fixed. The domain is no longer cached after a negative response to
1797 MAIL if the sender used is not empty.
1799 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1802 (c) A new callout option, "use_sender" has been added for people who want
1803 the modified behaviour.
1809 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1812 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1813 indeed breaks things for older releases.
1815 3. Added additional logging to the case where there is a problem reading data
1816 from a filter that is running in a subprocess using a pipe, in order to
1817 try to track down a specific problem.
1819 4. Testing facility fudge: when running in the test harness and attempting
1820 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1821 getting "No route to host". Convert this to a timeout.
1823 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1826 6. Some OS don't have socklen_t but use size_t instead. This affects the
1827 fifth argument of getsockopt() amongst other things. This is now
1828 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1829 can be set for individual OS. I have set it for SunOS5, OSF1, and
1830 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1831 some earlier ones do not.
1833 7. Change 4.30/15 was not doing the test caselessly.
1835 8. The standard form for an IPv6 address literal was being rejected by address
1836 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1837 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1838 this, as well as the form without the "IPv6" on the front (but only when
1839 address literals are enabled, of course).
1841 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1843 10. Exim crashed if a message with an empty sender address specified by -f
1844 encountered a router with an errors_to setting. This could be provoked only
1845 by a command such as
1849 where an empty string was supplied; "<>" did not hit this bug.
1851 11. Installed PCRE release 4.5.
1853 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1854 remained set. It is now erased.
1856 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1857 times from message ids (which are base 36 rather than the normal 62).
1859 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1860 were being counted as actual protocol errors, and logged if the log
1861 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1862 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1863 and DATA following a set of rejected RCPTs do not count as protocol errors.
1864 In other words, Exim assumes they were pipelined, though this may not
1865 actually be the case. Of course, in all cases the client gets an
1866 appropriate error code.
1868 15. If a lookup fails in an ACL condition, a message about the failure may
1869 be available; it is used if testing the ACL cannot continue, because most
1870 such messages specify what the cause of the deferral is. However, some
1871 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1872 that caused an old message to be retained and used if a later statement
1873 caused a defer, replacing the real cause of the deferral.
1875 16. If an IP address had so many PTR records that the DNS lookup buffer
1876 was not large enough to hold them, Exim could crash while trying to process
1877 the truncated data. It now detects and logs this case.
1879 17. Further to 4.21/58, another change has been made: if (and only if) the
1880 first line of a message (the first header line) ends with CRLF, a bare LF
1881 in a subsequent header line has a space inserted after it, so as not to
1882 terminate the header.
1884 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1885 unnecessarily, used atoi() instead of strtol(), and didn't check the
1886 termination when getting file sizes from file names by regex.
1888 19. Completely re-implemented the support for maildirsize files, in the light
1889 of a number of problems with the previous contributed implementation
1890 (4.30/29). In particular:
1892 . If the quota is zero, the maildirsize file is maintained, but no quota is
1895 . If the maildir directory does not exist, it is created before any attempt
1896 to write a maildirsize file.
1898 . The quota value in the file is just a cache; if the quota is changed in
1899 the transport, the new value overrides.
1901 . A regular expression is available for excluding directories from the
1904 20. The autoreply transport checks the characters in options that define the
1905 message's headers; it allows continued headers, but it was checking with
1906 isspace() after an embedded newline instead of explicitly looking for a
1909 21. If all the "regular" hosts to which an address was routed had passed their
1910 expiry times, and had not reached their retry times, the address was
1911 bounced, even if fallback hosts were defined. Now Exim should go on to try
1914 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1915 equivalent code in the SMTP transport. Some hosts send humungous responses
1916 to HELO/EHLO, more than 1024 it seems.
1918 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1919 gives compiler warnings in some environments. I've now done it "properly",
1922 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1923 (because of problems with the built-in one) was declared to return uschar *
1924 instead of char *, causing compiler failure.
1926 25. Fixed a file descriptor leak when processing alias/forward files.
1928 26. Fixed a minor format string issue in dbfn.c.
1930 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1932 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1933 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1936 29. When a sender address is verified, it is cached, to save repeating the test
1937 when there is more than one recipient in a message. However, when the
1938 verification involves a callout, it is possible for different callout
1939 options to be set for different recipients. It is too complicated to keep
1940 track of this in the cache, so now Exim always runs a verification when a
1941 callout is required, relying on the callout cache for the optimization.
1942 The overhead is duplication of the address routing, but this should not be
1945 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
1946 to be verified with callout=postmaster, and the main callout worked but the
1947 postmaster check failed, the verification correctly failed. However, if a
1948 subsequent RCPT command asked for sender verification *without* the
1949 postmaster check, incorrect caching caused this verification also to fail,
1952 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1953 it was not caching the DNS options (qualify_single, search_parents) that
1954 were used when the lookup failed. A subsequent lookup with different
1955 options therefore always gave the same answer, though there were cases
1956 where it should not have. (Example: a "domains = !$mx_any" option on a
1957 dnslookup router: the "domains" option is always processed without any
1958 widening, but the router might have qualify_single set.) Now Exim uses the
1959 cached value only when the same options are set.
1961 32. Added John Jetmore's "exipick" utility to the distribution.
1963 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
1964 than a timeout (e.g. a certificate is required, and is not provided), an
1965 Exim server now closes the connection immediately. Previously it waited for
1966 the client to close - but if the client is SSL, it seems that they each
1967 wait for each other, leading to a delay before one of them times out.
1969 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1970 maintained 0.8.x compatibility because I don't think many are using it, and
1971 it is clearly obsolete.
1973 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
1976 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
1977 client certificate was expired. A simple patch fixes this, though I don't
1978 understand the full logic of why the verify callback is called multiple
1981 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
1982 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
1983 which causes problems with some clients (such as the Certicom SSL Plus
1984 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
1985 disables the coutermeasure allowing Eudora to connect."
1987 38. Exim was not checking that a write() to a log file succeeded. This could
1988 lead to Bad Things if a log got too big, in particular if it hit a file
1989 size limit. Exim now panics and dies if it cannot write to a log file, just
1990 as it does if it cannot open a log file.
1992 39. Modified OS/Makefile-Linux so that it now contains
1994 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
1996 The two -D definitions ensure that Exim is compiled with large file
1997 support, which makes it possible to handle log files that are bigger than
2000 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2001 instance) a domain was checked against a named list that involved a lookup,
2002 causing $domain_data to be set, then another domain was checked against the
2003 same list, then the first domain was re-checked, the value of $domain_data
2004 after the final check could be wrong. In particular, if the second check
2005 failed, it could be set empty. This bug probably also applied to
2008 41. The strip_trailing_dot option was not being applied to the address given
2009 with the -f command-line option.
2011 42. The code for reading a message's header from the spool was incrementing
2012 $received_count, but never initializing it. This meant that the value was
2013 incorrect (doubled) while delivering a message in the same process in which
2014 it was received. In the most common configuration of Exim, this never
2015 happens - a fresh exec is done - but it can happen when
2016 deliver_drop_privilege is set.
2018 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2019 it now includes up to 150 characters of the unexpected data in the log
2022 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2023 and building data strings. The size of both of these buffers was 10 000
2024 bytes - far larger than anybody would *ever* want, thought I. Needless to
2025 say, somebody hit the limit. I have increased the maximum line length to
2026 20 000 and the maximum data length of concatenated lines to 100 000. I have
2027 also fixed two bugs, because there was no checking on these buffers. Tsk,
2028 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2029 buffer is too small.
2031 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2032 lsearch lookups. Now it does.
2034 46. When parsing a route_list item in a manualroute router, a fixed-length
2035 buffer was used for the list of hosts. I made this 1024 bytes long,
2036 thinking that nobody would ever have a list of hosts that long. Wrong.
2037 Somebody had a whole pile of complicated expansion conditions, and the
2038 string was silently truncated, leading to an expansion error. It turns out
2039 that it is easier to change to an unlimited length (owing to other changes
2040 that have happened since this code was originally written) than to build
2041 structure for giving a limitation error. The length of the item that
2042 expands into the list of hosts is now unlimited.
2044 47. The lsearch lookup could not handle data where the length of text line was
2045 more than 4095 characters. Such lines were truncated, leading to shortened
2046 data being returned. It should now handle lines of any length.
2048 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2049 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2052 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2053 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2054 no longer happens in comment lines. A list of replacements is now placed
2055 at the head of all of the source files, except those whose only change is
2056 to replace PERL_COMMAND in the very first #! line.
2058 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2059 messages on the queue, with a bottom-up merge sort, using code contributed
2060 by Michael Haardt. This should make operations like -bp somewhat faster on
2061 large queues. It won't affect queue runners, except when queue_run_in_order
2064 51. Installed eximstats 1.31 in the distribution.
2066 52. Added support for SRV lookups to the dnslookup router.
2068 53. If an ACL referred to $message_body or $message_body_end, the value was not
2069 reset for any messages that followed in the same SMTP session.
2071 54. The store-handling optimization for building very long strings was not
2072 differentiating between the different store pools. I don't think this
2073 actually made any difference in practice, but I've tidied it.
2075 55. While running the routers to verify a sender address, $sender_address
2076 was still set to the sender address. This is wrong, because when routing to
2077 send a bounce to the sender, it would be empty. Therefore, I have changed
2078 it so that, while verifying a sender address, $sender_address is set to <>.
2079 (There is no change to what happens when verifying a recipient address.)
2081 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2082 target A or AAAA records (if not already returned) without resetting the
2083 qualify_single or search_parents options of the DNS resolver. These are
2084 inappropriate in this case because the targets of MX and SRV records must
2085 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2086 target that, when qualified, matched something in the local domain. These
2087 two options are now turned off when doing these lookups.
2089 57. It seems that at least some releases of Reiserfs (which does not have the
2090 concept of a fixed number of inodes) returns zero and not -1 for the
2091 number of available inodes. This interacted badly with check_spool_inodes,
2092 which assumed that -1 was the "no such thing" setting. What I have done is
2093 to check that the total number of inodes is greater than zero before doing
2094 the test of how many are available.
2096 58. When a "warn" ACL statement has a log_message modifier, the message is
2097 remembered, and not repeated. This is to avoid a lot of repetition when a
2098 message has many recipients that cause the same warning to be written.
2099 Howewer, Exim was preserving the list of already written lines for an
2100 entire SMTP session, which doesn't seem right. The memory is now reset if a
2101 new message is started.
2103 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2104 debugging output unless log_rewrite was also set.
2106 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2107 of (int)(handle) when we know that handle contains (void *)(-1).
2109 61. The Exim daemon panic-logs an error return when it closes the incoming
2110 connection. However "connection reset by peer" seems to be common, and
2111 isn't really an error worthy of noting specially, so that particular error
2114 62. When Exim is trying to find all the local interfaces, it used to panic and
2115 die if the ioctl to get the interface flags failed. However, it seems that
2116 on at least one OS (Solaris 9) it is possible to have an interface that is
2117 included in the list of interfaces, but for which you get a failure error
2118 for this call. This happens when the interface is not "plumbed" into a
2119 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2120 failure of the "get flags" call assumes that the interface is down.
2122 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2123 makes life easier for people who are doing arithmetic on fields extracted
2124 from dates, where you often get leading zeros that should not be
2125 interpreted as octal.
2127 64. Added qualify_domain to the redirect router, to override the global
2130 65. If a pathologically long header line contained very many addresses (the
2131 report of this problem mentioned 10 000) and each of them was rewritten,
2132 Exim could use up a very large amount of memory. (It kept on making new
2133 copies of the header line as it rewrote, and never released the old ones.)
2134 At the expense of a bit more processing, the header rewriting function has
2135 been changed so that it no longer eats memory in this way.
2137 66. The generation of the Received: header has been moved from the time that a
2138 message starts to be received, to the time that it finishes. The timestamp
2139 in the Received: header should now be very close to that of the <= log
2140 line. There are two side-effects of this change:
2142 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2143 logged header lines no longer include the local Received: line, because
2144 it has not yet been created. The same applies to a copy of the message
2145 that is returned to a non-SMTP sender when a message is rejected.
2147 (b) When a filter file is tested using -bf, no additional Received: header
2148 is added to the test message. After some thought, I decided that this
2151 This change does not affect the value of $received_for. It is still set
2152 after address rewriting, but before local_scan() is called.
2154 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2156 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2157 gave an unhelpful panic error message, and a defer error. I have managed to
2158 change this behaviour so that it now rejects any supplied certificate,
2159 which seems right, as the list of acceptable certificates is empty.
2161 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2162 gave an unhelpful defer error. I have not managed to make this reject any
2163 supplied certificates, but the error message it gives is "no certificate
2164 supplied", which is not helpful.
2166 70. exigrep's output now also includes lines that are not associated with any
2167 message, but which match the given pattern. Implemented by a patch from
2168 Martin Sluka, which also tidied up the Perl a bit.
2170 71. Recipient callout verification, like sender verification, was using <> in
2171 the MAIL FROM command. This isn't really the right thing, since the actual
2172 sender may affect whether the remote host accepts the recipient or not. I
2173 have changed it to use the actual sender in the callout; this means that
2174 the cache record is now keyed on a recipient/sender pair, not just the
2175 recipient address. There doesn't seem to be a real danger of callout loops,
2176 since a callout by the remote host to check the sender would use <>.
2177 [SEE ABOVE: changed after hitting problems.]
2179 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2180 temporary errors. However, in the case of such a code being given after
2181 the end of a data transmission (i.e. after ".") Exim was failing to write
2182 a retry record for the message. (Yes, there was some broken host that was
2183 actually sending 8xx at this point.)
2185 73. An unknown lookup type in a host list could cause Exim to panic-die when
2186 the list was checked. (An example that provoked this was putting <; in the
2187 middle of a list instead of at the start.) If this happened during a DATA
2188 ACL check, a -D file could be left lying around. This kind of configuration
2189 error no longer causes Exim to die; instead it causes a defer errror. The
2190 incident is still logged to the main and panic logs.
2192 74. Buglet left over from Exim 3 conversion. The message "too many messages
2193 in one connection" was written to the rejectlog but not the mainlog, except
2194 when address rewriting (yes!) was being logged.
2196 75. Added write_rejectlog option.
2198 76. When a system filter was run not as root (that is, when system_filter_user
2199 was set), the values of the $n variables were not being returned to the
2200 main process; thus, they were not subsequently available in the $sn
2203 77. Added +return_path_on_delivery log selector.
2205 78. A connection timeout was being treated differently from recipients deferred
2206 when testing hosts_max_try with a message that was older than the host's
2207 retry timeout. (The host should not be counted, thus allowing all hosts to
2208 be tried at least once before bouncing.) This may have been the cause of an
2209 occasionally reported bug whereby a message would remain on the queue
2210 longer than the retry timeout, but would be bounced if a delivery was
2211 forced. I say "may" because I never totally pinned down the problem;
2212 setting up timeout/retry tests is difficult. See also the next item.
2214 79. The ultimate address timeout was not being applied to errors that involved
2215 a combination of host plus message (for example, a timeout on a MAIL
2216 command). When an address resolved to a number of possible hosts, and they
2217 were not all tried for each delivery (e.g. because of hosts_max_try), a
2218 message could remain on the queue longer than the retry timeout.
2220 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2223 81. Fixed an obscure SMTP outgoing bug which required at least the following
2224 conditions: (a) there was another message waiting for the same server;
2225 (b) the server returned 5xx to all RCPT commands in the first message so
2226 that the message was not completed; (c) the server dropped the connection
2227 or gave a negative response to the RSET that Exim sends to abort the
2228 transaction. The observed case was a dropped connection after DATA that had
2229 been sent in pipelining mode. That is, the server had advertised PIPELINING
2230 but was not implementing it correctly. The effect of the bug was incorrect
2231 behaviour, such as trying another host, and this could lead to a crash.
2237 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2238 and daemon.c were passed as pointers to ints; they should have been
2239 pointers to socklen_t variables (which are typically unsigned ints).
2241 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2244 3. Fixed a really odd bug that affected only the testing scheme; patching a
2245 certain fixed string in the binary changed the value of another string that
2246 happened to be identical to the end of the original first string.
2248 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2249 name", it returns that address as the IP address. On some operating
2250 systems (e.g. Solaris), it also passes back the IP address string as the
2251 "host name". However, on others (e.g. Linux), it passes back an empty
2252 string. Exim wasn't checking for this, and was changing the host name to an
2253 empty string, assuming it had been canonicized.
2255 5. Although rare, it is permitted to have more than one PTR record for a given
2256 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2257 all the names associated with an address, because they do in Solaris.
2258 However, it seems that they do not in Linux for data that comes from the
2259 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2260 I found this out when I moved to a new Linux workstation and tried to run
2261 the Exim test suite.
2263 To get round this problem I have changed the code so that it now does its
2264 own call to the DNS to look up PTR records when searching for a host name.
2265 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2266 addresses that are only in /etc/hosts are still found.
2268 This behaviour is, however, controlled by an option called host_lookup_
2269 order, which defaults to "bydns:byaddr". If people want to use the other
2270 order, or indeed, just use one or the other means of lookup, they can
2271 specify it in this variable.
2273 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2274 some operating systems, this comes back from gethostbyaddr() as an empty
2275 string, and this is what Exim used to test for. However, it seems that in
2276 other systems, "." is yielded. Exim now tests for this case too.
2278 7. The values of check_spool_space and check_log_space are now held internally
2279 as a number of kilobytes instead of an absolute number of bytes. If a
2280 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2281 kilobyte. This means that much larger values can be stored.
2283 8. Exim monitor: an attempt to get the action menu when not actually pointing
2284 at a message produces an empty menu entitled "No message selected". This
2285 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2286 no entries in it ("Shell widget menu has zero width and/or height"). So I
2287 have added a single, blank menu entry in this case.
2289 9. Added ${quote_local_part.
2291 10. MIME decoding is now applied to the contents of Subject: header lines when
2294 11. Now that a reference to $sender_host_address automatically causes a reverse
2295 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2296 host lookup before query-style lookups in lists that might use this
2297 variable. This has therefore been abolished, and the "net-" prefix is no
2298 longer necessary for query-style lookups.
2300 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2301 have been a typo for LFLAGS, so it has been changed.
2303 13. The install script calls Exim with "-C /dev/null" in order to find the
2304 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2305 to be output. Howeve, since Exim outputs its version number before the
2306 error, it didn't break the script. It just looked ugly. I fixed this by
2307 always allowing "-C /dev/null" if the caller is root.
2309 14. Ignore overlarge ACL variable number when reading spool file - insurance
2310 against a later release with more variables having written the file.
2312 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2313 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2314 this, as well as the form without the "IPv6" on the front.
2316 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2317 OS/Makefile-Darwin file.
2319 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2320 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2322 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2323 variable. However, some OS use ints for inodes. I've added cast to long int
2324 to get rid of the compiler warning.
2326 19. I had forgotten to lock out "/../" in configuration file names when
2327 ALT_CONFIG_PREFIX was set.
2329 20. Routers used for verification do not need to specify transports. However,
2330 if such a router generated a host list, and callout was configured, Exim
2331 crashed, because it could not find a port number from the (non-existent)
2332 transport. It now assumes port 25 in this circumstance.
2334 21. Added the -t option to exigrep.
2336 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2337 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2338 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2339 them is set, LOOKUP_LSEARCH is forced.
2341 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2343 24. Added sender and host information to the "rejected by local_scan()" log
2344 line; previously there was no indication of these.
2346 25. Added .include_if_exists.
2348 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2349 Linux. It turns out that not all file systems support this. Apparently some
2350 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2351 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2352 not supported on the file descriptor, is now ignored when Exim is trying to
2353 sync a directory. This applies only to Linux.
2355 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2357 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2358 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2361 29. Added support for maildirsize files from supplied patch (modified a bit).
2363 30. The use of :fail: followed by an empty string could lead Exim to respond to
2364 sender verification failures with (e.g.):
2366 550 Verification failed for <xxx>
2367 550 Sender verify failed
2369 where the first response line was missing the '-' that indicates it is not
2370 the final line of the response.
2372 31. The loop for finding the name of the user that called Exim had a hardwired
2373 limit of 10; it now uses the value of finduser_retries, which is used for
2374 all other user lookups.
2376 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2379 33. Exim was neglecting to zero errno before one call of strtol() when
2380 expanding a string and expecting an integer value. On some systems this
2381 resulted in spurious "integer overflow" errors. Also, it was casting the
2382 result into an int without checking.
2384 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2385 did not work. The code looks as if it has *never* worked, though it appears
2386 to have been documented since at least releast 1.62. I have made it work.
2388 35. The "timeout_DNS" error in retry rules, also documented since at least
2389 1.62, also never worked. As it isn't clear exactly what this means, and
2390 clearly it isn't a major issue, I have abolished the feature by treating it
2391 as "timeout", and writing a warning to the main and panic logs.
2393 36. The display of retry rules for -brt wasn't always showing the error code
2396 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2397 timeout_connect_A, timeout_connect_MX.
2399 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2400 to the empty sender.
2402 39. The daemon was not analysing the content of -oX till after it had closed
2403 stderr and disconnected from the controlling terminal. This meant that any
2404 syntax errors were only noted on the panic log, and the return code from
2405 the command was 0. By re-arranging the code a little, I've made the
2406 decoding happen first, so such errors now appear on stderr, and the return
2407 code is 1. However, the actual setting up of the sockets still happens in
2408 the disconnected process, so errors there are still only recorded on the
2411 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2412 connections (as happens on some IP stacks) was logged at start up time as
2413 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2414 it from "IPv6 and IPv4", which means that two separate sockets are being
2417 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2418 says whether AF_INET or AF_INET6 was passed as an argument.
2420 42. Exiwhat output was messed up when time zones were included in log
2423 43. Exiwhat now gives more information about the daemon's listening ports,
2424 and whether -tls-on-connect was used.
2426 44. The "port" option of the smtp transport is now expanded.
2428 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2429 silently ignored. Now an error message is written to the main and panic
2432 46. There's a new ACL modifier called "logwrite" which writes to a log file
2433 as soon as it is encountered.
2435 47. Added $local_user_uid and $local_user_gid at routing time.
2437 48. Exim crashed when trying to verify a sender address that was being
2440 49. Exim was recognizing only a space character after ".include". It now also
2441 recognizes a tab character.
2443 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2444 extracting the relevant information from the specification. The man page no
2445 longer contains scrambled data for the -d option, and I've added a section
2446 at the front about calling Exim under different names.
2448 51. Added "extra_headers" argument to the "mail" command in filter files.
2450 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2453 53. Installed eximstats 1.29.
2455 54. Added transport_filter_timeout as a generic transport option.
2457 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2458 Cc: header lines. This was required by RFC 822, but it not required by RFC
2461 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2462 incoming messages that did not have them. Now it does so only if the
2463 message originates locally, that is, if there is no associated remote host
2464 address. When Resent- header lines are present, this applies to the Resent-
2465 lines rather than the non-Resent- lines.
2467 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2468 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2470 58. Messages for configuration errors now include the name of the main
2471 configuration file - useful now that there may be more than one file in a
2472 list (.included file names were always shown).
2474 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2475 for those rare installations that do not start the daemon as root or run it
2476 setuid root. I've cut out the call to initgroups() if the daemon is not
2479 60. The Exim user and group can now be bound into the binary as text strings
2480 that are looked up at the start of Exim's processing.
2482 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2484 62. Added $mailstore_basename variable.
2486 63. Installed patch to sieve.c from Michael Haardt.
2488 64. When Exim failed to open the panic log after failing to open the main log,
2489 the original message it was trying to log was written to stderr and debug
2490 output, but if they were not available (the usual case in production), it
2491 was lost. Now it is written to syslog before the two lines that record the
2492 failures to open the logs.
2494 65. Users' Exim filters run in subprocesses under the user's uid. It is
2495 possible for a "deliver" command or an alias in a "personal" command to
2496 provoke an address rewrite. If logging of address rewriting is configured,
2497 this fails because the process is not running as root or exim. There may be
2498 a better way of dealing with this, but for the moment (because 4.30 needs
2499 to be released), I have disabled address rewrite logging when running a
2500 filter in a non-root, non-exim process.
2506 1. The buildconfig auxiliary program wasn't quoting the value set for
2507 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2508 not defined. This bug was masked in 4.22 by the effect that was fixed in
2511 2. Some messages that were rejected after a message id was allocated were
2512 shown as "incomplete" by exigrep. It no longer does this for messages that
2513 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2515 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2516 have allow_domain_literals set, the ID did not get logged in the <= line.
2517 Domain literals are now always recognized in Message-ID: header lines.
2519 4. The first argument for a ${extract expansion item is the key name or field
2520 number. Leading and trailing spaces in this item were not being ignored,
2521 causing some misleading effects.
2523 5. When deliver_drop_privilege was set, single queue runner processes started
2524 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2525 same command in the process it spins off) were not dropping privilege.
2527 6. When the daemon running as "exim" started a queue runner, it always
2528 re-executed Exim in the spun-off process. This is a waste of effort when
2529 deliver_drop_privilege is set. The new process now just calls the
2530 queue-runner function directly.
2536 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2539 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2540 ignored. Though the use of -oP was forcing the writing of a pid file, it
2541 was always written to the default place.
2543 3. If the message "no IP address found for host xxxx" is generated during
2544 incoming verification, it is now followed by identification of the incoming
2545 connection (so you can more easily find what provoked it).
2547 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2549 5. Added some features to "harden" Exim a bit more against certain attacks:
2551 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2552 be put in Local/Makefile. This is like the never_users runtime option,
2553 but it cannot be overridden. The default setting is "root".
2555 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2556 prefix string with which any file named in a -C command line option
2559 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2560 is retained for -C and -D only if the caller of Exim is root. Without
2561 it, the exim user may also use -C and -D and retain privilege.
2563 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2564 command line option is disabled.
2566 6. Macro names set by the -D option must start with an upper case letter, just
2567 like macro names defined in the configuration file.
2569 7. Added "dereference=" facility to LDAP.
2571 8. Two instances of the typo "uknown" in the source files are fixed.
2573 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2574 the Configure-Makefile script screwed up while processing it.
2576 10. Incorporated PCRE 4.4.
2578 11. The SMTP synchronization check was not operating right at the start of an
2579 SMTP session. For example, it could not catch a HELO sent before the client
2580 waited for the greeting. There is now a check for outstanding input at the
2581 point when the greeting is written. Because of the duplex, asynchronous
2582 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2583 way, but not yet received, when the check is performed.
2585 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2586 on TCP/IP sockets, because this apparently causes some broken clients to
2589 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2590 unchanged) from the Cygwin maintainer.
2592 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2593 was supported instead of testing for mbx. Effectively a typo.
2595 15. The spa authenticator server code was not checking that the input it
2596 received was valid base64.
2598 16. The debug output line for the "set" modifier in ACLs was not showing the
2599 name of the variable that was being set.
2601 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2603 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2604 lookup on its own. Something else was needed to trigger the lookup. For
2605 example, a match in host_lookup or the need for a host name in a host list.
2606 Now, if $sender_host_name is referenced and the host name has not yet been
2607 looked up, a lookup is performed. If the lookup fails, the variable remains
2608 empty, and $host_lookup_failed is set to "1".
2610 19. Added "eqi" as a case-independent comparison operator.
2612 20. The saslauthd authentication condition could segfault if neither service
2613 nor realm was specified.
2615 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2616 error message that was logged was misleading, and incoming SMTP
2617 connections were dropped. The message is now more accurate, and temporary
2618 errors are given to SMTP connections.
2620 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2621 (except RSET) with a 421 error, until QUIT is received. However, it was
2622 failing to send a response to QUIT.
2624 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2625 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2626 ACL is now run after the helo_try_verify_hosts code.
2628 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2629 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2630 case-independent, so other case variants are also recognized). Apparently
2631 some systems use these upper case variants.
2633 25. If more than two messages were waiting for the same host, and a transport
2634 filter was specified for the transport, Exim sent two messages over the
2635 same TCP/IP connection, and then failed with "socket operation on non-
2636 socket" when it tried to send the third.
2638 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2640 27. The extern definition of crypt16() in expand.c was not being excluded when
2641 the OS had its own crypt16() function.
2643 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2644 as a preferred synonym for return_size_limit, both as an option and as an
2647 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2649 30. Changed the default configuration ACL to relax the local part checking rule
2650 for addresses that are not in any local domains. For these addresses,
2651 slashes and pipe symbols are allowed within local parts, but the sequence
2652 /../ is explicitly forbidden.
2654 31. SPA server authentication was not clearing the challenge buffer before
2657 32. log_message in a "warn" ACL statement was writing to the reject log as
2658 well as to the main log, which contradicts the documentation and doesn't
2659 seem right (because no rejection is happening). So I have stopped it.
2661 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2662 However, I am unable to do any testing of this.
2664 34. Fixed an infelicity in the appendfile transport. When checking directories
2665 for a mailbox, to see if any needed to be created, it was accidentally
2666 using path names with one or more superfluous leading slashes; tracing
2667 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2669 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2670 contents are added to the log line that is written for every discarded
2671 recipient. (Previously a log_message setting was ignored.)
2673 36. The ${quote: operator now quotes the string if it is empty.
2675 37. The install script runs exim in order to find its version number. If for
2676 some reason other than non-existence or emptiness, which it checks, it
2677 could not run './exim', it was installing it with an empty version number,
2678 i.e. as "exim-". This error state is now caught, and the installation is
2681 38. An argument was missing from the function that creates an error message
2682 when Exim fails to connect to the socket for saslauthd authentication.
2683 This could cause Exim to crash, or give a corrupted message.
2685 39. Added isip, isip4, and isip6 to ${if conditions.
2687 40. The ACL variables $acl_xx are now saved with the message, and can be
2688 accessed later in routers, transports, and filters.
2690 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2691 strings in the file are not string-expanded.
2693 42. If a MAIL command specified a SIZE value that was too large to fit into an
2694 int variable, the check against message_size_limit failed. Such values are
2695 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2696 day this will have to be increased, but I don't think I want to be around
2697 when emails are that large.
2704 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2705 iconv() is not standard in FreeBSD.
2707 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2708 IPv6 enabled. The observed symptom was a segmentation fault on return from
2709 the function os_common_find_running_interfaces() in src/os.c.
2711 3. In the check_special_case() function in daemon.c I had used "errno" as an
2712 argument name, which causes warnings on some systems. This was basically a
2713 typo, since it was named "eno" in the comments!
2715 4. The code that waits for the clock to tick (at a resolution of some fraction
2716 of a second) so as to ensure message-id uniqueness was always waiting for
2717 at least one whole tick, when it could have waited for less. [This is
2718 almost certainly not relevant at current processor speeds, where it is
2719 unlikely to ever wait at all. But we try to future-proof.]
2721 5. The function that sleeps for a time interval that includes fractions of a
2722 second contained a race. It did not block SIGALRM between setting the
2723 timer, and suspending (a couple of lines later). If the interval was short
2724 and the sigsuspend() was delayed until after it had expired, the suspension
2725 never ended. On busy systems this could lead to processes getting stuck for
2728 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2729 process, before it forks any delivery processes. The open lookup caching
2730 mechanism meant that the open file or database connection was passed into
2731 the delivery process. The problem was that delivery processes always tidy
2732 up cached lookup data. This could cause a problem for the next delivery
2733 process started by the queue runner, because the external queue runner
2734 process does not know about the closure. So the next delivery process
2735 still has data in the lookup cache. In the case of a file lookup, there was
2736 no problem because closing a file descriptor in a subprocess doesn't affect
2737 the parent. However, if the lookup was caching a connection to a database,
2738 the connection was closed, and the second delivery process was likely to
2739 see errors such as "PGSQL: query failed: server closed the connection
2740 unexpectedly". The problem has been fixed by closing all cached lookups
2741 in a queue runner before running a delivery process.
2743 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2744 seem to have the "const" qualifier which it has on other OS. I've
2747 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2748 *of the same type* (client or server) with the same public name that an
2749 error should be diagnosed.
2751 9. When Exim looked up a host name for an IP address, but failed to find the
2752 original IP address when looking up the host name (a safety check), it
2753 output the message "<ip address> does not match any IP for NULL", which was
2754 confusing, to say the least. The bug was that the host name should have
2755 appeared instead of "NULL".
2757 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2758 user that is built into the binary, and the -C or -D options is used, root
2759 privilege is dropped before the configuration file is read. In addition,
2760 logging is switched to stderr instead of the normal log files. If the
2761 configuration then re-defines the Exim user, the unprivileged environment
2762 is probably not what is expected, so Exim logs a panic warning message (but
2765 However, if deliver_drop_privilege is set, the unprivileged state may well
2766 be exactly what is intended, so the warning has been cut out in that case,
2767 and Exim is allowed to try to write to its normal log files.
2773 1. smtp_return_error_details was not giving details for temporary sender
2774 or receiver verification errors.
2776 2. Diagnose a configuration error if two authenticators have the same public
2779 3. Exim used not to create the message log file for a message until the first
2780 delivery attempt. This could be confusing when incoming messages were held
2781 for policy or load reasons. The message log file is now created at the time
2782 the message is received, and an initial "Received" line is written to it.
2784 4. The automatically generated man page for command line options had a minor
2785 bug that caused no ill effects; however, a more serious problem was that
2786 the procedure for building the man page automatically didn't always
2787 operate. Consequently, release 4.20 contains an out-of-date version. This
2788 shouldn't happen again.
2790 5. When building Exim with embedded Perl support, the script that builds the
2791 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2792 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2794 6. The freeze_tell option was not being used for messages that were frozen on
2795 arrival, either by an ACL or by local_scan().
2797 7. Added the smtp_incomplete_transaction log selector.
2799 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2800 was accepting AUTH without a new EHLO.
2802 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2803 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2806 10. Logging of TCP/IP connections (when configured) now happens in the main
2807 daemon process instead of the child process, so that the TCP/IP connection
2808 count is more accurate (but it can never be perfect).
2810 11. The use of "drop" in a nested ACL was not being handled correctly in the
2811 outer ACL. Now, if condition failure induced by the nested "drop" causes
2812 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2813 or "require"), the connection is dropped.
2815 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2816 that yield "discard" can now be used with an "accept" or a "discard" verb,
2817 but an error is generated for any others (because I can't see a useful way
2818 to define what should happen).
2820 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2821 are now processed in the same way as lines in the Exim configuration file.
2822 In particular, continuation lines are supported.
2824 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2826 15. Added -ti meaning -t -i.
2828 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2829 domains, and warn by logging if others are found.
2831 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2832 returned from GIFCONF when Exim is trying to find the list of interfaces on
2833 a host. The code in os.c has been modified to copy each ifreq to an aligned
2834 structure in all cases.
2836 Also, in some cases, the returned ifreq's were being copied to a 'struct
2837 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2838 means the last couple of bytes of an IPv6 address could be chopped if the
2839 ifreq contained only a normal sockaddr (14 bytes storage).
2841 18. Named domain lists were not supported in the hosts_treat_as_local option.
2842 An entry such as +xxxx was not recognized, and was treated as a literal
2845 19. Ensure that header lines added by a DATA ACL are included in the reject log
2846 if the ACL subsequently rejects the message.
2848 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2849 MD5 (which is deprecated).
2851 21. When testing a filter file using -bf, Exim was writing a message when it
2852 took the sender from a "From " line in the message, but it was not doing so
2853 when it took $return_path from a Return-Path: header line. It now does.
2855 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2856 with a valid header line field name (a series of printing characters
2857 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2859 23. Changed "disc" in the source to "disk" to conform to the documentation and
2860 the book and for uniformity.
2862 24. Ignore Sendmail's -Ooption=value command line item.
2864 25. When execve() failed while trying to run a command in a pipe transport,
2865 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2866 could be confused with a return value of 69 from the command itself. This
2867 has been changed to 127, the value the shell returns if it is asked to run
2868 a non-existent command. The wording for the related log line suggests a
2869 non-existent command as the problem.
2871 26. If received_header_text expands to an empty string, do not add a Received:
2872 header line to the message. (Well, it adds a token one on the spool, but
2873 marks it "old" so that it doesn't get used or transmitted.)
2875 27. Installed eximstats 1.28 (addition of -nt option).
2877 28. There was no check for failure on the call to getsockname() in the daemon
2878 code. This can fail if there is a shortage of resources on the system, with
2879 ENOMEM, for example. A temporary error is now given on failure.
2881 29. Contrary to the C standard, it seems that in some environments, the
2882 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2883 program. Exim now does this explicitly; it affects the formatting of
2884 timestamps using strftime().
2886 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2887 complaints. I've now initialized all the variables, to avoid this.
2889 32. Header lines added by a system filter were not being "seen" during
2890 transport-time rewrites.
2892 33. The info_callback() function passed to OpenSSL is set up with type void
2893 (*)(SSL *, int, int), as described somewhere. However, when calling the
2894 function (actually a macro) that sets it up, the type void(*)() is
2895 expected. I've put in a cast to prevent warnings from picky compilers.
2897 34. If a DNS black list lookup found a CNAME record, but there were no A
2898 records associated with the domain it pointed at, Exim crashed.
2900 35. If a DNS black list lookup returned more than one A record, Exim ignored
2901 all but the first. It now scans all returned addresses if a particular IP
2902 value is being sought. In this situation, the contents of the
2903 $dnslist_value variable are a list of all the addresses, separated by a
2906 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2907 to accept a host name and all its aliases if the forward lookup for any of
2908 them yielded the IP address of the incoming connection. Now it accepts only
2909 those names whose forward lookup yields the correct IP address. Any other
2910 names are discarded. This closes a loophole whereby a rogue DNS
2911 administrator could create reverse DNS records to break through a
2912 wildcarded host restriction in an ACL.
2914 37. If a user filter or a system filter that ran in a subprocess used any of
2915 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2916 the wrong values were passed to the pipe command ($thisaddress had the
2917 value of $0, $0 had the value of $1, etc). This bug was introduced by
2918 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2920 38. Improved the line breaking for long SMTP error messages from ACLs.
2921 Previously, if there was no break point between 40 and 75 characters, Exim
2922 left the rest of the message alone. Two changes have been made: (a) I've
2923 reduced the minimum length to 35 characters; (b) if it can't find a break
2924 point between 35 and 75 characters, it looks ahead and uses the first one
2925 that it finds. This may give the occasional overlong line, but at least the
2926 remaining text gets split now.
2928 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2929 file descriptors might be low, and that setting 1000 would always raise it.
2930 It turns out that in some environments, the limit is already over 1000 and
2931 that lowering it causes trouble. So now Exim takes care not to decrease it.
2933 40. When delivering a message, the value of $return_path is set to $sender_
2934 address at the start of routing (routers may change the value). By an
2935 oversight, this default was not being set up when an address was tested by
2936 -bt or -bv, which affected the outcome if any router or filter referred to
2939 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2940 log only when "message" or "log_message" are set. However, if one of the
2941 conditions was an address verification, or a call to a nested ACL, the
2942 messages generated by the underlying test were being passed through. This
2943 no longer happens. The underlying message is available in $acl_verify_
2944 message for both "message" and "log_message" expansions, so it can be
2945 passed through if needed.
2947 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2948 new expansion $bh_ to give the encoded byte string without charset
2949 translation. Translation happens only if iconv() is available; HAVE_ICONV
2950 indicates this at build time. HEADERS_CHARSET gives the charset to
2951 translate to; headers_charset can change it in the configuration, and
2952 "headers charset" can change it in an individual filter file.
2954 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
2955 that creates RFC 2047 encoded "words" labels them as that charset instead
2956 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2957 expansion operator; (ii) when Exim creates a From: line for a local
2958 message; (iii) when a header line is rewritten to include a "phrase" part.
2960 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2961 buggy, causing it to skip the first lines of messages whose message ID
2962 ended in 'D'. This would not have bitten before Exim release 4.14, because
2963 message IDs were unlikely to end in 'D' before then. The effect was to have
2964 incorrect size information for certain domains.
2966 45. #include "config.h" was missing at the start of the crypt16.c module. This
2967 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2970 46. If there was a timeout during a "random" callout check, Exim treated it as
2971 a failure of the random address, and carried on sending RSET and the real
2972 address. If the delay was just some slowness somewhere, the response to the
2973 original RCPT would be taken as a response to RSET and so on, causing
2974 mayhem of various kinds.
2976 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
2977 when I implemented it. It didn't allow for the fact that some option values
2978 may legitimatetly be negative (e.g. size_addition), and it didn't even do
2979 the right test for positive values.
2981 48. Domain names in DNS records are case-independent. Exim always looks them up
2982 in lower case. Some resolvers return domain names in exactly the case they
2983 appear in the zone file, that is, they may contain uppercase letters. Not
2984 all resolvers do this - some return always lower case. Exim was treating a
2985 change of case by a resolver as a change of domain, similar to a widening
2986 of a domain abbreviation. This triggered its re-routing code and so it was
2987 trying to route what was effectively the same domain again. This normally
2988 caused routing to fail (because the router wouldn't handle the domain
2989 twice). Now Exim checks for this case specially, and just changes the
2990 casing of the domain that it ultimately uses when it transmits the message
2993 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
2996 50. If a filter generated a file delivery with a non-absolute name (possible if
2997 no home directory exists for the router), the forbid_file option was not
3000 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3001 the existing equality matching.
3003 52. Exim was using ints instead of ino_t variables in some places where it was
3004 dealing with inode numbers.
3006 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3007 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3008 called TMPDIR, and if it finds it is different, it changes its value.
3010 54. The smtp_printf() function is now made available to local_scan() so
3011 additional output lines can be written before returning. There is also an
3012 smtp_fflush() function to enable the detection of a dropped connection.
3013 The variables smtp_input and smtp_batched_input are exported to
3016 55. Changed the default runtime configuration: the message "Unknown user"
3017 has been removed from the ACL, and instead placed on the localuser router,
3018 using the cannot_route_message feature. This means that any verification
3019 failures that generate their own messages won't get overridden. Similarly,
3020 the "Unrouteable address" message that was in the ACL for unverifiable
3021 relay addresses has also been removed.
3023 56. Added hosts_avoid_esmtp to the smtp transport.
3025 57. The exicyclog script was not checking for the esoteric option
3026 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3027 will work only if exicyclog is run under the appropriate euid.
3029 58. Following a discussion on the list, the rules by which Exim recognises line
3030 endings on incoming messages have been changed. The -dropcr and drop_cr
3031 options are now no-ops, retained only for backwards compatibility. The
3032 following line terminators are recognized: LF CRLF CR. However, special
3033 processing applies to CR:
3035 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3036 nor a local message in the state where . is a terminator.
3038 (ii) If a bare CR is encountered in a header line, an extra space is added
3039 after the line terminator so as not to end the header. The reasoning
3040 behind this is that bare CRs in header lines are most likely either
3041 to be mistakes, or people trying to play silly games.
3043 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3044 was being incorrectly given as 18 bytes larger than it should have been.
3045 This is a VOB (very old bug).
3047 60. This may never have affected anything current, but just in case it has:
3048 When the local host is found other than at the start of a list of hosts,
3049 the local host, those with the same MX, and any that follow, are discarded.
3050 When the list in question was part of a longer list of hosts, the following
3051 hosts (not currently being processed) were also being discarded. This no
3052 longer happens. I'm not sure if this situation could ever has previously
3055 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3058 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3059 header. This is something that is recommended in a new Internet Draft, and
3060 is something that is documented as being done by Sendmail. There are two
3061 possible values. For messages generated by the autoreply transport, Exim
3064 Auto-Submitted: auto-replied
3066 whereas for all other generated messages (e.g. bounces) it adds
3068 Auto-Submitted: auto-generated
3070 63. The "personal" condition in filters now includes a test for the
3071 Auto-Submitted: header. If it contains the string "auto-" the message it
3072 not considered personal.
3074 64. Added rcpt_include_affixes as a generic transport option.
3076 65. Added queue_only_override (default true).
3078 66. Added the syslog_duplication option.
3080 67. If what should have been the first header line of a message consisted of
3081 a space followed by a colon, Exim was mis-interpreting it as a header line.
3082 It isn't of course - it is syntactically invalid and should therefore be
3083 treated as the start of the message body. The misbehaviour could have
3084 caused a number of strange effects, including loss of data in subsequent
3085 header lines, and spool format errors.
3087 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3088 client host had authenticated. This control can now be exercised by an ACL
3089 for more flexibility.
3091 69. By default, callouts do not happen when testing with -bh. There is now a
3092 variant, -bhc, which does actually run the callout code, including
3093 consulting and updating the callout cache.
3095 70. Added support for saslauthd authentication, courtesy of Alexander
3098 71. If statvfs() failed on the spool or log directories while checking their
3099 size for availability, Exim confusingly gave the error "space shortage".
3100 Furthermore, in debugging mode it crashed with a floating point exception.
3101 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3102 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3103 serious problem, Exim now writes to the main and panic logs when this
3104 happens, with details of the failure. It then refuses to accept the
3105 incoming message, giving the message "spool directory problem" or "log
3106 directory problem" with a 421 code for SMTP messages.
3108 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3109 0, 1, and 2 exist, because some OS complain for execs without them (see
3110 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3111 descriptors. However, the code omitted to check that the open succeeded,
3112 causing mysterious errors if for some reason the permissions on /dev/null
3113 got screwed. Now Exim writes a message to the main and panic logs, and
3114 bombs out if it can't open /dev/null.
3116 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3117 interact so that it is all more flexible. It is supposed to remain
3118 backwards compatible. Also added extra_local_interfaces.
3120 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3121 to bomb out with an assertion failure - to the client this appears as a
3122 connection drop. This problem occurs in the part of the code that was taken
3123 from the Samba project. Fortunately, the assertion is in a very simple
3124 function, so I have fixed this by reproducing the function inline in the
3125 one place where it is called, and arranging for authentication to fail
3126 instead of killing the process with assert().
3128 75. The SPA client code was not working when the server requested OEM rather
3129 than Unicode encoding.
3131 76. Added code to make require_files with a specific uid setting more usable in
3132 the case where statting the file as root fails - usually a non-root-mounted
3133 NFS file system. When this happens and the failure is EACCES, Exim now
3134 forks a subprocess and does the per-uid checking as the relevant uid.
3136 77. Added process_log_path.
3138 78. If log_file_path was not explicitly set, a setting of check_log_space or
3139 check_log_inodes was ignored.
3141 79. If a space check for the spool or log partitions fails, the incident is now
3142 logged. Of course, in the latter case the data may get lost...
3144 80. Added the %p formatting code to string_format() so that it can be used to
3145 print addresses in debug_print(). Adjusted all the address printing in the
3146 debugging in store.c to use %p rather than %d.
3148 81. There was a concern that a line of code in smtp_in.c could overflow a
3149 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3150 initially expressed, the concern was not well-founded, because trailing
3151 spaces are removed early. However, if the trailing spaces were followed by
3152 a NULL, they did not get removed, so the overflow was possible. Two fixes
3155 (a) I re-wrote the offending code in a cleaner fashion.
3156 (b) If an incoming SMTP command contains a NULL character, it is rejected
3159 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3160 runs initgroups(), so that if the Exim user is in any additional groups,
3161 they will be used during message reception.
3167 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob