1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.192 2005/08/02 09:24:45 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Added the "success_on_redirect" address verification option. See
10 NewStuff for rationale and an example.
12 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
14 PH/02 Patch to exigrep to allow it to work on syslog lines.
16 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
17 fread() to skip over the body file's header line, because in Cygwin the
18 header line is locked and is inaccessible.
20 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
21 co-exist for some time) to make it clear that it is the Exim ID that is
22 referenced, not the Message-ID: header line.
24 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
25 string_format() function, because snprintf() does not exist on all
28 PH/06 The use of forbid_filter_existstest now also locks out the use of the
29 ${stat: expansion item.
31 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
32 protocol synchronization error", to keep the pedants happy.
34 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
35 well as for IRIX systems, when gcc is being used. See the host.c source
38 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
40 PH/10 Named domain lists were not working if used in a queue_smtp_domains
47 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
49 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
50 command itself ended in error, the underlying message about the transport
51 filter timeout was being overwritten with the pipe command error. Now the
52 underlying error message should be appended to the second error message.
54 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
56 PH/02 Reduce the amount of output that "make" produces by default. Full output
57 can still be requested.
59 PH/03 The warning log line about a condition test deferring for a "warn" verb
60 was being output only once per connection, rather than after each
61 occurrence (because it was using the same function as for successful
62 "warn" verbs). This seems wrong, so I have changed it.
64 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
65 it should not have, which might have caused a crash in the right
66 circumstances, but probably never did.
68 PH/04 Installed a modified version of Tony Finch's patch to make submission
69 mode fix the return path as well as the Sender: header line, and to
70 add a /name= option so that you can make the user's friendly name appear
73 TF/03 Added the control = fakedefer ACL modifier.
75 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
76 Mark Lowes for thorough testing.
78 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
80 TK/03 Merged latest SRS patch from Miles Wilton.
82 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
83 with the definition in sysexits.h (which is #included earlier).
84 Fortunately, Exim does not actually use EX_OK. The code used to try to
85 preserve the sysexits.h value, by assumimg that macro definitions were
86 scanned for macro replacements. I have been disabused of this notion,
87 so now the code just undefines EX_OK before #including unistd.h.
89 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
90 in the smtp transport. When a block could not be written in a single
91 write() function, the timeout was being re-applied to each part-write.
92 This seems wrong - if the receiver was accepting one byte at a time it
93 would take for ever. The timeout is now adjusted when this happens. It
94 doesn't have to be particularly precise.
96 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
97 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
99 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
100 without a domain if the check to <postmaster@domain> fails.
102 SC/01 Eximstats: added -xls and the ability to specify output files
103 (patch written by Frank Heydlauf).
105 SC/02 Eximstats: use FileHandles for outputing results.
107 SC/03 Eximstats: allow any combination of xls, txt, and html output.
109 SC/04 Eximstats: fixed display of large numbers with -nvr option
111 SC/05 Eximstats: fixed merging of reports with empty tables.
113 SC/06 Eximstats: added the -include_original_destination flag
115 SC/07 Eximstats: removed tabs and trailing whitespace.
117 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
119 TK/06 MBOX spool code: Add real "From " MBOX separator line
120 so the .eml file is really in mbox format (even though
121 most programs do not really care). Patch from Alex Miller.
123 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
124 The latter is generated from $received_to and is only set if the
125 message has one envelope recipient. SA can use these headers,
126 obviously out-of-the-box. Patch from Alex Miller.
128 PH/08 The ${def test on a variable was returning false if the variable's
129 value was "0", contrary to what the specification has always said!
130 The result should be true unless the variable is empty.
132 PH/09 The syntax error of a character other than { following "${if
133 def:variable_name" (after optional whitespace) was not being diagnosed.
134 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
135 accidental colon was present, for example, could give incorrect results.
137 PH/10 Tidied the code in a number of places where the st_size field of a stat()
138 result is used (not including appendfile, where other changes are about
141 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
142 This involved changing a lot of size variables from int to off_t. It
143 should work with maildirs and everything.
145 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
146 spamd dying while we are connected to it.
148 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
151 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
152 function for BDB 4.3.
154 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
156 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
157 the sources for PCRE have been re-arranged and more files are now
160 PH/15 The code I had for printing potentially long long variables in PH/11
161 above was not the best (it lost precision). The length of off_t variables
162 is now inspected at build time, and an appropriate printing format (%ld
163 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
164 to be "long long int" or "long int". This is needed for the internal
165 formatting function string_vformat().
167 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
168 the configuration file to be ":syslog", then the script "guesses" where
169 the logs files are, rather than using the compiled in default. In our
170 case the guess is not the same as the compiled default, so the script
171 suddenly stopped working when I started to use syslog. The patch checks
172 to see if log_file_path is "". If so, it attempts to read it from exim
173 with no configuration file to get the compiled in version, before it
174 falls back to the previous guessing code."
176 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
177 implementing BATV in an Exim configuration. See NewStuff for the gory
180 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
181 Makefile that are specific to HP-UX.
183 PH/18 If the "use_postmaster" option was set for a recipient callout together
184 with the "random" option, the postmaster address was used as the MAIL
185 FROM address for the random test, but not for the subsequent recipient
186 test. It is now used for both.
188 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
189 patch removes a few documentation additions to RFC 3028, because the
190 latest draft now contains them. It adds the new en;ascii-case comparator
191 and a new error check for 8bit text in MIME parts. Comparator and
192 require names are now matched exactly. I enabled the subaddress
193 extension, but it is not well tested yet (read: it works for me)."
195 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
196 rework some of the code of TK/09 above to avoid the hardwired use of
197 "%lld" and "long long". Replaced the call to snprintf() with a call to
200 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
201 records point to non-existent hosts", "retry timeout exceeded", and
202 "retry time not reached for any host after a long failure period".
204 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
205 experimental DomainKeys support:
207 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
208 (2) On an error such as an illegally used "control", the wrong name for
209 the control was given.
211 These problems did NOT occur unless DomainKeys support was compiled.
213 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
215 PH/24 Added ${if match_ip condition.
217 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
218 databases so that it will be absolutely obvious if a crash occurs in the
219 DB library. This is a regular occurrence (often caused by mis-matched
222 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
223 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
224 picked up on a user's system that detects such things. There doesn't seem
225 to be a gcc warning option for this - only an attribute that has to be
226 put on the function's prototype. It seems that in Fedora Core 4 they have
227 set this on a number of new functions. No doubt there will be more in due
230 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
231 specify a transport. However, if an address that was verified by such a
232 router was the subject of a callout, Exim crashed because it tried to
233 read the rcpt_include_affixes from the non-existent transport. Now it
234 just assumes that the setting of that option is false. This bug was
235 introduced by 4.51/PH/31.
237 PH/28 Changed -d+all to exclude +memory, because that information is very
238 rarely of interest, but it makes the output a lot bigger. People tend to
239 do -d+all out of habit.
241 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
242 code in os-type was giving problems when libc.so lives in lib64, like on
245 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
246 aren't the modern standard, and it seems that some systems' include files
247 don't always have them. Exim was already checking for some of the newer
248 ones like T_AAAA, and defining it itself. I've added checks for all the
249 record types that Exim uses.
251 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
252 not automatically generating a new one, as it is supposed to. This
253 prevented TLS from working. If the file did exist, but contained invalid
254 data, a new version was generated, as expected. It was only the case of a
255 non-existent file that was broken.
257 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
258 with a change in libdomainkeys > 0.64.
260 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
261 from DNS. If the selector record carries the flag, it now has
262 precedence over the domain-wide flag.
264 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
266 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
267 the use of a port name that isn't defined in /etc/services) occurred, the
268 message was deferred as in a normal delivery, and thus remained on the
269 spool, instead of being failed because of the mua_wrapper setting. This
270 is now fixed, and I tidied up some of the mua_wrapper messages at the
273 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
274 the messages in a hash of arrays rather than using individual hashes.
275 This is a bit cleaner and results in dramatic memory savings, albeit
276 at a slight CPU cost.
278 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
279 as requested by Marc Sherman.
281 SC/10 Eximstats: added histograms for user specified patterns as requested
284 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
286 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
287 fopen() in the content-scanning modules that did not already have it.
293 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
294 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
296 TK/02 Fix ACL "control" statement not being available in MIME ACL.
298 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
300 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
301 to test Sieve filters that use "vacation".
303 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
304 that changes the way the GnuTLS parameters are stored in the cache file.
305 The new format can be generated externally. For backward compatibility,
306 if the data in the cache doesn't make sense, Exim assumes it has read an
307 old-format file, and it generates new data and writes a new file. This
308 means that you can't go back to an older release without removing the
311 PH/03 A redirect router that has both "unseen" and "one_time" set does not
312 work if there are any delivery delays because "one_time" forces the
313 parent to be marked "delivered", so its unseen clone is never tried
314 again. For this reason, Exim now forbids the simultaneous setting of
317 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
318 redirected to themselves ("homonym" addresses). Read the long ChangeLog
319 entry if you want to know the details. The fix, however, neglected to
320 consider the case when local delivery batching is involved. The test for
321 "previously delivered" was not happening when checking to see if an
322 address could be batched with a previous (undelivered) one; under
323 certain circumstances this could lead to multiple deliveries to the same
326 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
327 in its include files, and this causes problems building Exim.
329 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
330 header_syntax) but Exim was just ignoring anything given after a slash.
331 In particular, this caused confusion with an attempt to use "verify =
332 reverse_host_lookup/defer_ok". An error is now given when options are
333 supplied for verify items that do not have them. (Maybe reverse_host_
334 lookup should have a defer_ok option, but that's a different point.)
336 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
337 defined by RFC 821) to 2048, because there were problems with some AUTH
338 commands, and RFC 1869 says the size should be increased for extended
339 SMTP commands that take arguments.
341 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
344 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
345 "unknown" error; now it says that the functionality isn't in the binary.
347 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
348 an address' error message when a string expansion fails (syntax or
349 whatever). Otherwise the password may appear in the log. Following change
350 PH/42 below, there is no longer a chance of it appearing in a bounce
353 PH/11 Installed exipick version 20050225.0 from John Jetmore.
355 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
356 of its addresses was ever tried. (Bugzilla bug #2.)
358 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
359 the result incorrectly in the debug output. (It correctly added a newline
360 to what was transported.)
362 TF/01 Added $received_time.
364 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
365 commented out examples of how to interface to a virus scanner and to
366 SpamAssassin. Also added commented examples of av_scanner and
367 spamd_address settings.
369 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
370 and controls are allowed in which ACLs. There were a couple of minor
371 errors. Some of the entries in the conditions table (which is a table of
372 where they are NOT allowed) were getting very unwieldy; rewrote them as a
373 negation of where the condition IS allowed.
375 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
377 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
378 header file does not have a version number, so I've had to invent a new
379 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
380 API. The code is untested by me (my Linux distribution still has 0.3.2 of
381 radiusclient), but it was contributed by a Radius user.
383 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
384 files or directories, for OpenSSL.
386 PH/19 When an Exim process that is running as root has to create an Exim log
387 file, it does so in a subprocess that runs as exim:exim so as to get the
388 ownership right at creation (otherwise, other Exim processes might see
389 the file with the wrong ownership). There was no test for failure of this
390 fork() call, which would lead to the process getting stuck as it waited
391 for a non-existent subprocess. Forks do occasionally fail when resources
392 run out. I reviewed all the other calls to fork(); they all seem to check
395 PH/20 When checking for unexpected SMTP input at connect time (before writing
396 the banner), Exim was not dealing correctly with a non-positive return
397 from the read() function. If the client had disconnected by this time,
398 the result was a log entry for a synchronization error with an empty
399 string after "input=" when read() returned zero. If read() returned -1
400 (an event I could not check), uninitialized data bytes were printed.
401 There were reports of junk text (parts of files, etc) appearing after
404 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
406 PH/22 Added support for macro redefinition, and (re)definition in between
407 driver and ACL definitions.
409 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
410 forgetting to use the resulting value; it was using the unexpanded value.
412 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
413 hadn't been configured. The fix is from Juergen Kreileder, who
414 understands it better than I do:
416 "Here's what I see happening with three configured cyrus_sasl
417 authenticators configured (plain, login, cram-md5):
419 On startup auth_cyrus_sasl_init() gets called for each of these.
420 This means three calls to sasl_listmech() without a specified mech_list.
421 => SASL tests which mechs of all available mechs actually work
422 => three warnings about OTP not working
423 => the returned list contains: plain, login, cram-md5, digest-md5, ...
425 With the patch, sasl_listmech() also gets called three times. But now
426 SASL's mech_list option is set to the server_mech specified in the the
427 authenticator. Or in other words, the answer from sasl_listmech()
428 gets limited to just the mech you're testing for (which is different
430 => the return list contains just 'plain' or 'login', 'cram-md5' or
431 nothing depending on the value of ob->server_mech.
433 I've just tested the patch: Authentication still works fine,
434 unavailable mechs specified in the exim configuration are still
435 caught, and the auth.log warnings about OTP are gone."
437 PH/25 When debugging is enabled, the contents of the command line are added
438 to the debugging output, even when log_selector=+arguments is not
441 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
442 answer is "GNU", and only if the return is "GNU/something" is the answer
445 PH/27 $acl_verify_message is now set immediately after the failure of a
446 verification in an ACL, and so is available in subsequent modifiers. In
447 particular, the message can be preserved by coding like this:
449 warn !verify = sender
450 set acl_m0 = $acl_verify_message
452 Previously, $acl_verify_message was set only while expanding "message"
453 and "log_message" when a very denied access.
455 PH/28 Modified OS/os.c-Linux with
457 -#ifndef OS_LOAD_AVERAGE
458 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
460 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
461 nomenclature these days.)
463 PH/29 Installed patch from the Sieve maintainer that adds the options
464 sieve_useraddress and sieve_subaddress to the redirect router.
466 PH/30 In these circumstances:
467 . Two addresses routed to the same list of hosts;
468 . First host does not offer TLS;
469 . First host accepts first address;
470 . First host gives temporary error to second address;
471 . Second host offers TLS and a TLS session is established;
472 . Second host accepts second address.
473 Exim incorrectly logged both deliveries with the TLS parameters (cipher
474 and peerdn, if requested) that were in fact used only for the second
477 PH/31 When doing a callout as part of verifying an address, Exim was not paying
478 attention to any local part prefix or suffix that was matched by the
479 router that accepted the address. It now behaves in the same way as it
480 does for delivery: the affixes are removed from the local part unless
481 rcpt_include_affixes is set on the transport.
483 PH/32 Add the sender address, as F=<...>, to the log line when logging a
484 timeout during the DATA phase of an incoming message.
486 PH/33 Sieve envelope tests were broken for match types other than :is. I have
487 applied a patch sanctioned by the Sieve maintainer.
489 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
490 the uid or gid is negative. A case of a negative gid caused this to be
491 noticed. The fix allows for either to be negative.
493 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
494 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
497 PH/36 The change PH/12 above was broken. Fixed it.
499 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
500 the grounds that routing the same address twice would always produce the
501 same answer. This might have been true once, but it is certainly no
502 longer true now. Routing a child address may depend on the previous
503 routing that produced that child. Some complicated redirection strategies
504 went wrong when messages had multiple recipients, and made Exim's
505 behaviour dependent on the order in which the addresses were given.
507 I have moved the duplicate checking until after the routing is complete.
508 Exim scans the addresses that are assigned to local and remote
509 transports, and removes any duplicates. This means that more work will be
510 done, as duplicates will always all be routed, but duplicates are
511 presumably rare, so I don't expect this is of any significance.
513 For deliveries to pipes, files, and autoreplies, the duplicate checking
514 still happens during the routing process, since they are not going to be
517 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
518 It corrects a timeout issue with spamd. This is Ian's comment: "The
519 background is that sometimes spamd either never reads data from a
520 connection it has accepted, or it never writes response data. The exiscan
521 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
522 blindly assumes that writes won't block so it may never time out."
524 PH/39 Allow G after quota size as well as K and M.
526 PH/40 The value set for $authenticated_id in an authenticator may not contain
527 binary zeroes or newlines because the value is written to log lines and
528 to spool files. There was no check on this. Now the value is run through
529 the string_printing() function so that such characters are converted to
530 printable escape sequences.
532 PH/41 $message_linecount is a new variable that contains the total number of
533 lines in the message. Compare $body_linecount, which is the count for the
536 PH/42 Exim no longer gives details of delivery errors for specific addresses in
537 bounce and delay warning messages, except in certain special cases, which
540 (a) An SMTP error message from a remote host;
541 (b) A message specified in a :fail: redirection;
542 (c) A message specified in a "fail" command in a system filter;
543 (d) A message specified in a FAIL return from the queryprogram router;
544 (e) A message specified by the cannot_route_message router option.
546 In these cases only, Exim does include the error details in bounce and
547 warning messages. There are also a few cases where bland messages such
548 as "unrouteable address" or "local delivery error" are given.
550 PH/43 $value is now also set for the "else" part of a ${run expansion.
552 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
553 being worked on, but at least Exim now implements the latest version to
556 PH/45 In a pipe transport, although a timeout while waiting for the pipe
557 process to complete was treated as a delivery failure, a timeout while
558 writing the message to the pipe was logged, but erroneously treated as a
559 successful delivery. Such timeouts include transport filter timeouts. For
560 consistency with the overall process timeout, these timeouts are now
561 treated as errors, giving rise to delivery failures by default. However,
562 there is now a new Boolean option for the pipe transport called
563 timeout_defer, which, if set TRUE, converts the failures into defers for
564 both kinds of timeout. A transport filter timeout is now identified in
567 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
568 systems where "make" and "gmake" are different, calling "gmake" at top
569 level broke things. I've arranged for the value of $(MAKE) to be passed
570 from the Makefile to this script so that it can call the same version of
574 A note about Exim versions 4.44 and 4.50
575 ----------------------------------------
577 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
578 changes of various kinds. As a consequence, a big documentation update was
579 needed. This delayed the release for rather longer than seemed good, especially
580 in the light of a couple of (minor) security issues. Therefore, the changes
581 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
582 release. So 4.44 and 4.50 are in effect two different branches that both start
585 I have left the 4.50 change log unchanged; it contains all the changes since
586 4.43. The change log for 4.44 is below; many of its items are identical to
587 those for 4.50. This seems to be the most sensible way to preserve the
588 historical information.
594 1. Minor wording change to the doc/README.SIEVE file.
596 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
597 computation of the current number of files was incorrect.
599 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
600 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
603 4. Give more explanation in the error message when the command for a transport
604 filter fails to execute.
606 5. There are several places where Exim runs a non-Exim command in a
607 subprocess. The SIGUSR1 signal should be disabled for these processes. This
608 was being done only for the command run by the queryprogram router. It is
609 now done for all such subprocesses. The other cases are: ${run, transport
610 filters, and the commands run by the lmtp and pipe transports.
612 6. Added CONFIGURE_GROUP build-time option.
614 7. Some older OS have a limit of 256 on the maximum number of file
615 descriptors. Exim was using setrlimit() to set 1000 as a large value
616 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
617 systems. I've change it so that if it can't get 1000, it tries for 256.
619 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
620 was an oversight, and furthermore, ever since the addition of extra
621 controls (e.g. 4.43/32), the checks on when to allow different forms of
622 "control" were broken. There should now be diagnostics for all cases when a
623 control that does not make sense is encountered.
625 9. Added the /retain_sender option to "control=submission".
627 10. $recipients is now available in the predata ACL (oversight).
629 11. Tidy the search cache before the fork to do a delivery from a message
630 received from the command line. Otherwise the child will trigger a lookup
631 failure and thereby defer the delivery if it tries to use (for example) a
632 cached ldap connection that the parent has called unbind on.
634 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
635 of $address_data from the recipient verification was clobbered by the
638 13. The value of address_data from a sender verification is now available in
639 $sender_address_data in subsequent conditions in the ACL statement.
641 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
643 15. Added a new option "connect=<time>" to callout options, to set a different
646 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
647 was its contents. (It was OK if the option was not defined at all.)
649 17. A "Completed" log line is now written for messages that are removed from
650 the spool by the -Mrm option.
652 18. New variables $sender_verify_failure and $recipient_verify_failure contain
653 information about exactly what failed.
655 19. Added -dd to debug only the daemon process.
657 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
658 handles timeouts, both on the server side and network timeouts. Renamed the
659 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
661 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
662 It is now set to "smtps".
664 22. $host_address is now set to the target address during the checking of
667 23. When checking ignore_target_hosts for an ipliteral router, no host name was
668 being passed; this would have caused $sender_host_name to have been used if
669 matching the list had actually called for a host name (not very likely,
670 since this list is usually IP addresses). A host name is now passed as
673 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
674 code that specifies a non-restarting handler (typically sigaction() in
675 modern systems) in an attempt to fix a rare and obscure crash bug.
677 25. Narrowed the window for a race in the daemon that could cause it to ignore
678 SIGCHLD signals. This is not a major problem, because they are used only to
679 wake it up if nothing else does.
681 26. A malformed maildirsize file could cause Exim to calculate negative values
682 for the mailbox size or file count. Odd effects could occur as a result.
683 The maildirsize information is now recalculated if the size or filecount
686 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
687 support for a long time. Removed HAVE_SYS_VFS_H.
689 28. Installed the latest version of exipick from John Jetmore.
691 29. In an address list, if the pattern was not a regular expression, an empty
692 subject address (from a bounce message) matched only if the pattern was an
693 empty string. Non-empty patterns were not even tested. This was the wrong
694 because it is perfectly reasonable to use an empty address as part of a
695 database query. An empty address is now tested by patterns that are
696 lookups. However, all the other forms of pattern expect the subject to
697 contain a local part and a domain, and therefore, for them, an empty
698 address still always fails if the pattern is not itself empty.
700 30. Exim went into a mad DNS loop when attempting to do a callout where the
701 host was specified on an smtp transport, and looking it up yielded more
704 31. Re-factored the code for checking spool and log partition space into a
705 function that finds that data and another that does the check. The former
706 is then used to implement four new variables: $spool_space, $log_space,
707 $spool_inodes, and $log_inodes.
709 32. The RFC2047 encoding function was originally intended for short strings
710 such as real names; it was not keeping to the 75-character limit for
711 encoded words that the RFC imposes. It now respects the limit, and
712 generates multiple encoded words if necessary. To be on the safe side, I
713 have increased the buffer size for the ${rfc2047: expansion operator from
716 33. It is now permitted to omit both strings after an "if" condition; if the
717 condition is true, the result is "true". As before, when the second string
718 is omitted, a false condition yields an empty string. This makes it less
719 cumbersome to write custom ACL and router conditions.
721 34. Failure to deliver a bounce message always caused it to be frozen, even if
722 there was an errors_to setting on the router. The errors_to setting is now
725 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
726 canonical form (fully expanded) before being placed in
727 $sender_host_address.
729 36. The table in the code that translates DNS record types into text (T_A to
730 "A" for instance) was missing entries for NS and CNAME. It is just possible
731 that this could have caused confusion if both these types were looked up
732 for the same domain, because the text type is used as part of Exim's
733 per-process caching. But the chance of anyone hitting this buglet seems
736 37. The dnsdb lookup has been extended in a number of ways.
738 (1) There is a new type, "zns", which walks up the domain tree until it
739 finds some nameserver records. It should be used with care.
741 (2) There is a new type, "mxh", which is like "mx" except that it returns
742 just the host names, not the priorities.
744 (3) It is now possible to give a list of domains (or IP addresses) to be
745 looked up. The behaviour when one of the lookups defers can be
746 controlled by a keyword.
748 (4) It is now possible to specify the separator character for use when
749 multiple records are returned.
751 38. The dnslists ACL condition has been extended: it is now possible to supply
752 a list of IP addresses and/or domains to be looked up in a particular DNS
755 39. Added log_selector=+queue_time_overall.
757 40. When running the queue in the test harness, wait just a tad after forking a
758 delivery process, to get repeatability of debugging output.
760 41. Include certificate and key file names in error message when GnuTLS fails
761 to set them up, because the GnuTLS error message doesn't include the name
762 of the failing file when there is a problem reading it.
764 42. Allow both -bf and -bF in the same test run.
766 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
768 44. The "Exiscan patch" is now merged into the mainline Exim source.
770 45. Sometimes the final signoff response after QUIT could fail to get
771 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
772 < 0 before doing a fflush(). This bug looks as though it goes back to the
773 introduction of TLS in release 3.20, but "sometimes" must have been rare
774 because the tests only now provoked it.
776 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
777 (this can affect the format of dates).
779 47. exim_tidydb, when checking for the continued existence of a message for
780 which it has found a message-specific retry record, was not finding
781 messages that were in split spool directories. Consequently, it was
782 deleting retry records that should have stayed in existence.
784 48. Steve fixed some bugs in eximstats.
786 49. The SPA authentication driver was not abandoning authentication and moving
787 on to the next authenticator when an expansion was forced to fail,
788 contradicting the general specification for all authenticators. Instead it
789 was generating a temporary error. It now behaves as specified.
791 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
792 (the order specifies the preference for clients). The order is now AES256,
793 AES128, 3DES, ARCFOUR128.
795 51. Small patch to Sieve code - explicitly set From: when generating an
798 52. Exim crashed if a remote delivery caused a very long error message to be
799 recorded - for instance if somebody sent an entire SpamAssassin report back
800 as a large number of 550 error lines. This bug was coincidentally fixed by
801 increasing the size of one of Exim's internal buffers (big_buffer) that
802 happened as part of the Exiscan merge. However, to be on the safe side, I
803 have made the code more robust (and fixed the comments that describe what
806 53. Now that there can be additional text after "Completed" in log lines (if
807 the queue_time_overall log selector is set), a one-byte patch to exigrep
808 was needed to allow it to recognize "Completed" as not the last thing in
811 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
812 patch that reportedly fixes this has been added. I am not expert enough to
813 create a test for it. This is what the patch creator wrote:
815 "I found a little strange behaviour of ldap code when working with
816 Windows 2003 AD Domain, where users was placed in more than one
817 Organization Units. When I tried to give exim partial DN, the exit code
818 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
819 But simultaneously result of request was absolutely normal ldap result,
820 so I produce this patch..."
822 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
823 REFERENCE, so I have modified the code to exclude the patch when that macro
826 55. Some experimental protocols are using DNS PTR records for new purposes. The
827 keys for these records are domain names, not reversed IP addresses. The
828 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
829 leaves it alone. Component reversal etc. now happens only for IP addresses.
832 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
834 57. Double the size of the debug message buffer (to 2048) so that more of very
835 long debug lines gets shown.
837 58. The exicyclog utility now does better if the number of log files to keep
838 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
840 59. Two changes related to the smtp_active_hostname option:
842 (1) $smtp_active_hostname is now available as a variable.
843 (2) The default for smtp_banner uses $smtp_active_hostname instead
844 of $primary_hostname.
846 60. The host_aton() function is supposed to be passed a string that is known
847 to be a valid IP address. However, in the case of IPv6 addresses, it was
848 not checking this. This is a hostage to fortune. Exim now panics and dies
849 if the condition is not met. A case was found where this could be provoked
850 from a dnsdb PTR lookup with an IPv6 address that had more than 8
851 components; fortuitously, this particular loophole had already been fixed
852 by change 4.50/55 above.
854 If there are any other similar loopholes, the new check in host_aton()
855 itself should stop them being exploited. The report I received stated that
856 data on the command line could provoke the exploit when Exim was running as
857 exim, but did not say which command line option was involved. All I could
858 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
862 61. There was a buffer overflow vulnerability in the SPA authentication code
863 (which came originally from the Samba project). I have added a test to the
864 spa_base64_to_bits() function which I hope fixes it.
867 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
868 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
870 63. The daemon start-up calls getloadavg() while still root for those OS that
871 need the first call to be done as root, but it missed one case: when
872 deliver_queue_load_max is set with deliver_drop_privilege. This is
873 necessary for the benefit of the queue runner, because there is no re-exec
874 when deliver_drop_privilege is set.
876 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
879 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
880 was in use, was not putting the data itself into the right store pool;
881 consequently, it could be overwritten for a subsequent message in the same
882 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
885 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
887 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
888 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
889 the function were treating the return as a boolean value, which happened to
890 work because 0=false and not-0=true, but is not correct code.
892 68. The host_aton() function was not handling scoped IPv6 addresses (those
893 with, for example, "%eth0" on the end) correctly.
895 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
896 negated items (that is, ~something) in unsigned ints. Some compilers
897 apparently mutter when there is no cast.
899 70. If an address verification called from an ACL failed, and did not produce a
900 user-specific message (i.e. there was only a "system" message), nothing was
901 put in $acl_verify_message. In this situation, it now puts the system
904 71. Change 4.23/11 added synchronization checking at the start of an SMTP
905 session; change 4.31/43 added the unwanted input to the log line - except
906 that it did not do this in the start of session case. It now does.
908 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
909 This is wrong and can cause the other end to generate a synchronization
910 error if it is another Exim or anything else that does the synchronization
911 check. A QUIT command is no longer sent after a timeout.
913 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
916 74. The defer_ok option of callout verification was not working if it was used
917 when verifying addresses in header lines, that is, for this case:
919 verify = header_sender/callout=defer_ok
921 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
922 those file descriptors could be used for SMTP connections. If anything
923 wrote to stderr (the example that came up was "warn" in embedded Perl), it
924 could be sent to the SMTP client, causing chaos. The daemon now opens
925 stdin, stdout, and stderr to /dev/null when it puts itself into the
928 76. Arrange for output from Perl's "warn" command to be written to Exim's main
929 log by default. The user can override this with suitable Perl magic.
931 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
932 the log message when discard triggers, was not working for the DATA ACL or
933 for the non-SMTP ACL.
935 78. Error message wording change in sieve.c.
937 79. If smtp_accept_max_per_host was set, the number of connections could be
938 restricted to fewer than expected, because the daemon was trying to set up
939 a new connection before checking whether the processes handling previous
940 connections had finished. The check for completed processes is now done
941 earlier. On busy systems, this bug wouldn't be noticed because something
942 else would have woken the daemon, and it would have reaped the completed
945 80. If a message was submitted locally by a user whose login name contained one
946 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
947 It caused a spool format error. I have fixed the spool reading code. A
948 related problem was that the "from" clause in the Received: line became
949 illegal because of the space(s). It is now covered by ${quote_local_part.
951 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
954 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
956 83. Patch from Sieve maintainer to fix unterminated string problem in
959 84. Some minor changes to the Linux configuration files to help with other
960 OS variants using glibc.
962 85. One more patch for Sieve to update vacation handling to latest spec.
965 ----------------------------------------------------
966 See the note above about the 4.44 and 4.50 releases.
967 ----------------------------------------------------
973 1. Change 4.43/35 introduced a bug that caused file counts to be
974 incorrectly computed when quota_filecount was set in an appendfile
977 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
978 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
981 3. Give more explanation in the error message when the command for a transport
982 filter fails to execute.
984 4. There are several places where Exim runs a non-Exim command in a
985 subprocess. The SIGUSR1 signal should be disabled for these processes. This
986 was being done only for the command run by the queryprogram router. It is
987 now done for all such subprocesses. The other cases are: ${run, transport
988 filters, and the commands run by the lmtp and pipe transports.
990 5. Some older OS have a limit of 256 on the maximum number of file
991 descriptors. Exim was using setrlimit() to set 1000 as a large value
992 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
993 systems. I've change it so that if it can't get 1000, it tries for 256.
995 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
996 was an oversight, and furthermore, ever since the addition of extra
997 controls (e.g. 4.43/32), the checks on when to allow different forms of
998 "control" were broken. There should now be diagnostics for all cases when a
999 control that does not make sense is encountered.
1001 7. $recipients is now available in the predata ACL (oversight).
1003 8. Tidy the search cache before the fork to do a delivery from a message
1004 received from the command line. Otherwise the child will trigger a lookup
1005 failure and thereby defer the delivery if it tries to use (for example) a
1006 cached ldap connection that the parent has called unbind on.
1008 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1009 of $address_data from the recipient verification was clobbered by the
1010 sender verification.
1012 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1013 was its contents. (It was OK if the option was not defined at all.)
1015 11. A "Completed" log line is now written for messages that are removed from
1016 the spool by the -Mrm option.
1018 12. $host_address is now set to the target address during the checking of
1019 ignore_target_hosts.
1021 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1022 being passed; this would have caused $sender_host_name to have been used if
1023 matching the list had actually called for a host name (not very likely,
1024 since this list is usually IP addresses). A host name is now passed as
1027 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1028 code that specifies a non-restarting handler (typically sigaction() in
1029 modern systems) in an attempt to fix a rare and obscure crash bug.
1031 15. Narrowed the window for a race in the daemon that could cause it to ignore
1032 SIGCHLD signals. This is not a major problem, because they are used only to
1033 wake it up if nothing else does.
1035 16. A malformed maildirsize file could cause Exim to calculate negative values
1036 for the mailbox size or file count. Odd effects could occur as a result.
1037 The maildirsize information is now recalculated if the size or filecount
1040 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1041 support for a long time. Removed HAVE_SYS_VFS_H.
1043 18. Updated exipick to current release from John Jetmore.
1045 19. Allow an empty sender to be matched against a lookup in an address list.
1046 Previously the only cases considered were a regular expression, or an
1049 20. Exim went into a mad DNS lookup loop when doing a callout where the
1050 host was specified on the transport, if the DNS lookup yielded more than
1053 21. The RFC2047 encoding function was originally intended for short strings
1054 such as real names; it was not keeping to the 75-character limit for
1055 encoded words that the RFC imposes. It now respects the limit, and
1056 generates multiple encoded words if necessary. To be on the safe side, I
1057 have increased the buffer size for the ${rfc2047: expansion operator from
1060 22. Failure to deliver a bounce message always caused it to be frozen, even if
1061 there was an errors_to setting on the router. The errors_to setting is now
1064 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1065 canonical form (fully expanded) before being placed in
1066 $sender_host_address.
1068 24. Updated eximstats to version 1.33
1070 25. Include certificate and key file names in error message when GnuTLS fails
1071 to set them up, because the GnuTLS error message doesn't include the name
1072 of the failing file when there is a problem reading it.
1074 26. Expand error message when OpenSSL has problems setting up cert/key files.
1077 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1078 (this can affect the format of dates).
1080 28. exim_tidydb, when checking for the continued existence of a message for
1081 which it has found a message-specific retry record, was not finding
1082 messages that were in split spool directories. Consequently, it was
1083 deleting retry records that should have stayed in existence.
1085 29. eximstats updated to version 1.35
1086 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1087 1.35 - bugfix such that pie charts by volume are generated correctly
1089 30. The SPA authentication driver was not abandoning authentication and moving
1090 on to the next authenticator when an expansion was forced to fail,
1091 contradicting the general specification for all authenticators. Instead it
1092 was generating a temporary error. It now behaves as specified.
1094 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1095 (the order specifies the preference for clients). The order is now AES256,
1096 AES128, 3DES, ARCFOUR128.
1098 31. Small patch to Sieve code - explicitly set From: when generating an
1101 32. Exim crashed if a remote delivery caused a very long error message to be
1102 recorded - for instance if somebody sent an entire SpamAssassin report back
1103 as a large number of 550 error lines. This bug was coincidentally fixed by
1104 increasing the size of one of Exim's internal buffers (big_buffer) that
1105 happened as part of the Exiscan merge. However, to be on the safe side, I
1106 have made the code more robust (and fixed the comments that describe what
1109 33. Some experimental protocols are using DNS PTR records for new purposes. The
1110 keys for these records are domain names, not reversed IP addresses. The
1111 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1112 leaves it alone. Component reversal etc. now happens only for IP addresses.
1115 34. The host_aton() function is supposed to be passed a string that is known
1116 to be a valid IP address. However, in the case of IPv6 addresses, it was
1117 not checking this. This is a hostage to fortune. Exim now panics and dies
1118 if the condition is not met. A case was found where this could be provoked
1119 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1120 components; fortuitously, this particular loophole had already been fixed
1121 by change 4.50/55 or 4.44/33 above.
1123 If there are any other similar loopholes, the new check in host_aton()
1124 itself should stop them being exploited. The report I received stated that
1125 data on the command line could provoke the exploit when Exim was running as
1126 exim, but did not say which command line option was involved. All I could
1127 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1128 running as the user.
1131 35. There was a buffer overflow vulnerability in the SPA authentication code
1132 (which came originally from the Samba project). I have added a test to the
1133 spa_base64_to_bits() function which I hope fixes it.
1136 36. The daemon start-up calls getloadavg() while still root for those OS that
1137 need the first call to be done as root, but it missed one case: when
1138 deliver_queue_load_max is set with deliver_drop_privilege. This is
1139 necessary for the benefit of the queue runner, because there is no re-exec
1140 when deliver_drop_privilege is set.
1142 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1143 was in use, was not putting the data itself into the right store pool;
1144 consequently, it could be overwritten for a subsequent message in the same
1145 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1148 38. Sometimes the final signoff response after QUIT could fail to get
1149 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1150 < 0 before doing a fflush(). This bug looks as though it goes back to the
1151 introduction of TLS in release 3.20, but "sometimes" must have been rare
1152 because the tests only now provoked it.
1158 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1159 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1160 Now it returns an integer. A number of calls were still expecting a T/F
1161 return. Fortuitously, in all cases, the tests worked in OK situations,
1162 which is the norm. However, things would have gone wrong on any write
1163 failures on the smtp file descriptor. This function is used when sending
1164 messages over SMTP and also when doing verify callouts.
1166 2. When Exim is called to do synchronous delivery of a locally submitted
1167 message (the -odf or -odi options), it no longer closes stderr before doing
1170 3. Implemented the mua_wrapper option.
1172 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1174 5. Implemented the functions header_remove(), header_testname(),
1175 header_add_at_position(), and receive_remove_recipient(), and exported them
1178 6. If an ACL "warn" statement specified the addition of headers, Exim already
1179 inserted X-ACL-Warn: at the start if there was no header name. However, it
1180 was not making this test for the second and subsequent header lines if
1181 there were newlines in the string. This meant that an invalid header could
1182 be inserted if Exim was badly configured.
1184 7. Allow an ACL "warn" statement to add header lines at the start or after all
1185 the Received: headers, as well as at the end.
1187 8. Added the rcpt_4xx retry error code.
1189 9. Added postmaster_mailfrom=xxx to callout verification option.
1191 10. Added mailfrom=xxxx to the callout verification option, for verify=
1194 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1195 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1197 12. Inserted some casts to stop certain compilers warning when using pointer
1198 differences as field lengths or precisions in printf-type calls (mostly
1199 affecting debugging statements).
1201 13. Added optional readline() support for -be (dynamically loaded).
1203 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1204 same clock tick as a message's arrival, so that its received time was the
1205 same as the "first fail" time on the retry record, and that message
1206 remained on the queue past the ultimate address timeout, every queue runner
1207 would try a delivery (because it was past the ultimate address timeout) but
1208 after another failure, the ultimate address timeout, which should have then
1209 bounced the address, did not kick in. This was a "< instead of <=" error;
1210 in most cases the first failure would have been in the next clock tick
1211 after the received time, and all would be well.
1213 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1214 being recognized when the domain list was tested by the match_domain
1215 condition in an expansion string.
1217 16. Added the ${str2b64: operator.
1219 17. Exim was always calling setrlimit() to set a large limit for the number of
1220 processes, without checking whether the existing limit was already
1221 adequate. (It did check for the limit on file descriptors.) Furthermore,
1222 errors from getrlimit() and setrlimit() were being ignored. Now they are
1223 logged to the main and panic logs, but Exim does carry on, to try to do its
1224 job under whatever limits there are.
1226 18. Imported PCRE 5.0.
1228 19. Trivial typo in log message " temporarily refused connection" (the leading
1231 20. If the log selector return_path_on_delivery was set and an address was
1232 redirected to /dev/null, the delivery process crashed because it assumed
1233 that a return path would always be set for a "successful" delivery. In this
1234 case, the whole delivery is bypassed as an optimization, and therefore no
1237 21. Internal re-arrangement: the function for sending a challenge and reading
1238 a response while authentication was assuming a zero-terminated challenge
1239 string. It's now changed to take a pointer and a length, to allow for
1240 binary data in such strings.
1242 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1244 23. Exim was not respecting finduser_retries when seeking the login of the
1245 uid under which it was called; it was always trying 10 times. (The default
1246 setting of finduser_retries is zero.) Also, it was sleeping after the final
1247 failure, which is pointless.
1249 24. Implemented tls_on_connect_ports.
1251 25. Implemented acl_smtp_predata.
1253 26. If the domain in control=submission is set empty, Exim assumes that the
1254 authenticated id is a complete email address when it generates From: or
1255 Sender: header lines.
1257 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1258 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1259 chown and chgrp in /bin and hostname in /usr/bin.
1261 28. Exim was keeping the "process log" file open after each use, just as it
1262 does for the main log. This opens the possibility of it remaining open for
1263 long periods when the USR1 signal hits a daemon. Occasional processlog
1264 errors were reported, that could have been caused by this. Anyway, it seems
1265 much more sensible not to leave this file open at all, so that is what now
1268 29. The long-running daemon process does not normally write to the log once it
1269 has entered its main loop, and it closes the log before doing so. This is
1270 so that log files can straightforwardly be renamed and moved. However,
1271 there are a couple of unusual error situations where the daemon does write
1272 log entries, and I had neglected to close the log afterwards.
1274 30. The text of an SMTP error response that was received during a remote
1275 delivery was being truncated at 512 bytes. This is too short for some of
1276 the long messages that one sometimes sees. I've increased the limit to
1279 31. It is now possible to make retry rules that apply only when a message has a
1280 specific sender, in particular, an empty sender.
1282 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1283 it possible to be selective about when SMTP synchronization is enforced.
1285 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1287 32. Implemented hosts_connection_nolog.
1289 33. Added an ACL for QUIT.
1291 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1294 35. Added mailbox_size and mailbox_filecount to appendfile.
1296 36. Added control = no_multiline_responses to ACLs.
1298 37. There was a bug in the logic of the code that waits for the clock to tick
1299 in the case where the clock went backwards by a substantial amount such
1300 that the microsecond fraction of "now" was more than the microsecond
1301 fraction of "then" (but the whole seconds number was less).
1303 38. Added support for the libradius Radius client library this is found on
1304 FreeBSD (previously only the radiusclient library was supported).
1310 1. When certain lookups returned multiple values in the form name=value, the
1311 quoting of the values was not always being done properly. Specifically:
1312 (a) If the value started with a double quote, but contained no whitespace,
1314 (b) If the value contained whitespace other than a space character (i.e.
1315 tabs or newlines or carriage returns) it was not quoted.
1316 This fix has been applied to the mysql and pgsql lookups by writing a
1317 separate quoting function and calling it from the lookup code. The fix
1318 should probably also be applied to nisplus, ibase and oracle lookups, but
1319 since I cannot test any of those, I have not disturbed their existing code.
1321 2. A hit in the callout cache for a specific address caused a log line with no
1322 reason for rejecting RCPT. Now it says "Previous (cached) callout
1323 verification failure".
1325 3. There was an off-by-one bug in the queryprogram router. An over-long
1326 return line was truncated at 256 instead of 255 characters, thereby
1327 overflowing its buffer with the terminating zero. As well as fixing this, I
1328 have increased the buffer size to 1024 (and made a note to document this).
1330 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1331 when Exim is waiting for an SMTP response from a remote server, Exim
1332 restarts its select() call on the socket, thereby resetting its timeout.
1333 This is not a problem when such interrupts are rare. Somebody set up a cron
1334 job to run exiwhat every 2 minutes, which is less than the normal select()
1335 timeout (5 or 10 minutes). This meant that the select() timeout never
1336 kicked in because it was always reset. I have fixed this by comparing the
1337 time when an interrupt arrives with the time at the start of the first call
1338 to select(). If more time than the timeout has elapsed, the interrupt is
1339 treated as a timeout.
1341 5. Some internal re-factoring in preparation for the addition of Sieve
1342 extensions (by MH). In particular, the "personal" test is moved to a
1343 separate function, and given an option for scanning Cc: and Bcc: (which is
1344 not set for Exim filters).
1346 6. When Exim created an email address using the login of the caller as the
1347 local part (e.g. when creating a From: or Sender: header line), it was not
1348 quoting the local part when it contained special characters such as @.
1350 7. Installed new OpenBSD configuration files.
1352 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1353 try to make them clearer.
1355 9. Callout options, other than the timeout value, were being ignored when
1356 verifying sender addresses in header lines. For example, when using
1358 verify = header_sender/callout=no_cache
1360 the cache was (incorrectly) being used.
1362 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1363 only the Cygwin environment.
1365 11. When return_path_on_delivery was set as a log selector, if different remote
1366 addresses in the same message used different return paths and parallel
1367 remote delivery occurred, the wrong values would sometimes be logged.
1368 (Whenever a remote delivery process finished, the return path value from
1369 the most recently started remote delivery process was logged.)
1371 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1372 lines when AUTH and/or TLS are in use. This is the "received protocol"
1373 field. Exim used to use "asmtp" for authenticated SMTP, without any
1374 indication (in the protocol name) for TLS use. Now it follows the RFC and
1375 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1376 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1377 names appear in log lines as well as in Received: header lines.
1379 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1380 extensions, and comparison tests, and to fix some bugs.
1382 14. Changes to the "personal" filter test:
1384 (1) The test was buggy in that it was just doing the equivalent of
1385 "contains" tests on header lines. For example, if a user's address was
1386 anne@some.where, the "personal" test would incorrectly be true for
1388 To: susanne@some.where
1390 This test is now done by extracting each address from the header in turn,
1391 and checking the entire address. Other tests that are part of "personal"
1392 are now done using regular expressions (for example, to check local parts
1393 of addresses in From: header lines).
1395 (2) The list of non-personal local parts in From: addresses has been
1396 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1397 taken from the Sieve specification recommendations.
1399 (3) If the message contains any header line starting with "List-" it is
1400 treated as non-personal.
1402 (4) The test for "circular" in the Subject: header line has been removed
1403 because it now seems ill-conceived.
1405 15. Minor typos in src/EDITME comments corrected.
1407 16. Installed latest exipick from John Jetmore.
1409 17. If headers_add on a router specified a text string that was too long for
1410 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1411 of string_sprintf() is now avoided.
1413 18. $message_body_size was not set (it was always zero) when running the DATA
1414 ACL and the local_scan() function.
1416 19. For the "mail" command in an Exim filter, no default was being set for
1417 the once_repeat time, causing a random time value to be used if "once" was
1418 specified. (If the value happened to be <= 0, no repeat happened.) The
1419 default is now 0s, meaning "never repeat". The "vacation" command was OK
1420 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1421 (I found it when inspecting the code).
1423 20. There is now an overall timeout for performing a callout verification. It
1424 defaults to 4 times the callout timeout, which applies to individual SMTP
1425 commands during the callout. The overall timeout applies when there is more
1426 than one host that can be tried. The timeout is checked before trying the
1427 next host. This prevents very long delays if there are a large number of
1428 hosts and all are timing out (e.g. when the network connections are timing
1429 out). The value of the overall timeout can be changed by specifying an
1430 additional sub-option for "callout", called "maxwait". For example:
1432 verify = sender/callout=5s,maxwait=20s
1434 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1435 to the end before writing, but this should make it even safer).
1437 22. Exim was forgetting that it had advertised PIPELINING for the second and
1438 subsequent messages on an SMTP connection. It was also not resetting its
1439 memory on STARTTLS and an internal HELO.
1441 23. When Exim logs an SMTP synchronization error within a session, it now
1442 records whether PIPELINING has been advertised or not.
1444 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1445 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1446 rather than long int.
1448 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1450 26. Added the never_mail option to autoreply.
1456 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1457 crash if the getsockname() call failed; this can happen if a connection is
1458 closed very soon after it is established. The problem was simply in the
1459 order in which certain operations were done, causing Exim to try to write
1460 to the SMTP stream before it had set up the file descriptor. The bug has
1461 been fixed by making things happen in the correct order.
1467 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1468 before the connection was closed, thus losing the rejection response.
1470 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1471 some early Solaris releases, but causes trouble in current releases where
1472 socklen_t is defined.
1474 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1477 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1480 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1481 that releasing the top of it at the end releases what was used for sub-
1482 expansions (unless the block got too big). However, discard this block if
1483 the first thing is a variable or header, so that we can use its block when
1484 it is dynamic (useful for very large $message_headers, for example).
1486 6. Lookups now cache *every* query, not just the most recent. A new, separate
1487 store pool is used for this. It can be recovered when all lookup caches are
1488 flushed. Lookups now release memory at the end of their result strings.
1489 This has involved some general refactoring of the lookup sources.
1491 7. Some code has been added to the store_xxx() functions to reduce the amount
1492 of flapping under certain conditions.
1494 8. log_incoming_interface used to affect only the <= reception log lines. Now
1495 it causes the local interface and port to be added to several more SMTP log
1496 lines, for example "SMTP connection from", and rejection lines.
1498 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1500 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1502 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1503 could be overwritten at the end of the current message (or the start of a
1504 new message if it was set in a HELO ACL). The value is now preserved for
1505 the duration of the SMTP connection.
1507 12. If a transport had a headers_rewrite setting, and a matching header line
1508 contained an unqualified address, that address was qualified, even if it
1509 did not match any rewriting rules. The underlying bug was that the values
1510 of the flags that permit the existence of unqualified sender and recipient
1511 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1512 non-local messages, and by -bnq for local messages) were not being
1513 preserved with the message after it was received.
1515 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1516 "next input=" as part of the text comprising the host identity instead of
1517 the correct text. The code was using the same buffer for two different
1518 strings. However, depending on which order the printing function evaluated
1519 its arguments, the bug did not always show up. Under Linux, for example, my
1520 test suite worked just fine.
1522 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1523 doesn't work with some older versions of Perl. It has been changed to "my",
1524 which in any case is probably the better facility to use.
1526 15. A really picky compiler found some instances of statements for creating
1527 error messages that either had too many or two few arguments for the format
1530 16. The size of the buffer for calls to the DNS resolver has been increased
1531 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1532 for addresses that have a lot of PTR records. This alleviates a problem; it
1533 does not fully solve it.
1535 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1536 buffer now truncates the list and logs the incident, which is the same
1537 action as happens when Exim is looking up a host name and its aliases.
1538 Previously in this situation something unpredictable would happen;
1539 sometimes it was "internal error: store_reset failed".
1541 18. If a server dropped the connection unexpectedly when an Exim client was
1542 using GnuTLS and trying to read a response, the client delivery process
1543 crashed while trying to generate an error log message.
1545 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1546 string, for example:
1548 warn message = H1: something\nH2: something
1550 the text was added as a single header line from Exim's point of view
1551 though it ended up OK in the delivered message. However, searching for the
1552 second and subsequent header lines using $h_h2: did not work. This has been
1553 fixed. Similarly, if a system filter added multiple headers in this way,
1554 the routers could not see them.
1556 20. Expanded the error message when iplsearch is called with an invalid key to
1557 suggest using net-iplsearch in a host list.
1559 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1560 ACLs are no longer actually imposed (and a message to that effect is
1563 22. If a "gecos" field in a passwd entry contained escaped characters, in
1564 particular, if it contained a \" sequence, Exim got it wrong when building
1565 a From: or a Sender: header from that name. A second bug also caused
1566 incorrect handling when an unquoted " was present following a character
1567 that needed quoting.
1569 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1570 was not being matched caselessly.
1572 24. Arranged for all hyphens in the exim.8 source to be escaped with
1575 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1576 properly. Recipient callout cache records were still being keyed to include
1577 the sender, even when use_sender was set false. This led to far more
1578 callouts that were necessary. The sender is no longer included in the key
1579 when use_sender is false.
1581 26. Added "control = submission" modifier to ACLs.
1583 27. Added the ${base62d: operator to decode base 62 numbers.
1585 28. dnsdb lookups can now access SRV records.
1587 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1588 the configuration file.
1590 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1591 (-v) mode. This makes the output for a verbose queue run more intelligible.
1593 31. Added a use_postmaster feature to recipient callouts.
1595 32. Added the $body_zerocount variable, containing the number of binary zero
1596 bytes in the message body.
1598 33. The time of last modification of the "new" subdirectory is now used as the
1599 "mailbox time last read" when there is a quota error for a maildir
1602 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1604 35. Added +ignore_unknown as a special item in host lists.
1606 36. Code for decoding IPv6 addresses in host lists is now included, even if
1607 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1608 address was recognized as an IP address, but was then not correctly decoded
1609 into binary, causing unexpected and incorrect effects when compared with
1616 1. Very minor rewording of debugging text in manualroute to say "list of
1617 hosts" instead of "hostlist".
1619 2. If verify=header_syntax was set, and a header line with an unqualified
1620 address (no domain) and a large number of spaces between the end of the
1621 name and the colon was received, the reception process suffered a buffer
1622 overflow, and (when I tested it) crashed. This was caused by some obsolete
1623 code that should have been removed. The fix is to remove it!
1625 3. When running in the test harness, delay a bit after writing a bounce
1626 message to get a bit more predictability in the log output.
1628 4. Added a call to search_tidyup() just before forking a reception process. In
1629 theory, someone could use a lookup in the expansion of smtp_accept_max_
1630 per_host which, without the tidyup, could leave open a database connection.
1632 5. Added the variables $recipient_data and $sender_data which get set from a
1633 lookup success in an ACL "recipients" or "senders" condition, or a router
1634 "senders" option, similar to $domain_data and $local_part_data.
1636 6. Moved the writing of debug_print from before to after the "senders" test
1639 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1640 problems for message scanning, either using a data ACL, or using
1641 local_scan() because the Received: header was not generated till after they
1642 were called (in order to set the time as the time of reception completion).
1643 I have revised the way this works. The header is now generated after the
1644 body is received, but before the ACL or local_scan() are called. After they
1645 are run, the timestamp in the header is updated.
1651 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1652 before starting a queue runner without re-exec. This happened only when
1653 deliver_drop_privilege was set or when the Exim user was set to root. The
1654 effect of the bug was that timeouts during subsequent deliveries caused
1655 crashes instead of being properly handled. The handler is now left at its
1656 default (and expected) setting.
1658 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1659 message, again when deliver_drop_privilege is set or Exim is run as root.
1660 The bug described in (1) was not present in this case, but the tidying up
1661 of the other signals was missing. I have made the two cases consistent.
1663 3. The ignore_target_hosts setting on a manualroute router was being ignored
1664 for hosts that were looked up using the /MX notation.
1666 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1669 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1670 operated on the sender address. After changing the $sender_address to <>
1671 for the sender address verify, Exim was re-instated it as the original
1672 (before rewriting) address, but remembering that it had rewritten it, so it
1673 wasn't rewriting it again. This bug also had the effect of breaking the
1674 sender address verification caching when the sender address was rewritten.
1676 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1677 This has been changed so that if the ip literal address matches
1678 ignore_target_hosts, the router declines.
1680 7. Added expansion conditions match_domain, match_address, and match_local_
1681 part (NOT match_host).
1683 8. The placeholder for the Received: header didn't have a length field set.
1685 9. Added code to Exim itself and to exim_lock to test for a specific race
1686 condition that could lead to file corruption when using MBX delivery. The
1687 issue is with the lockfile that is created in /tmp. If this file is removed
1688 after a process has opened it but before that process has acquired a lock,
1689 there is the potential for a second process to recreate the file and also
1690 acquire a lock. This could lead to two Exim processes writing to the file
1691 at the same time. The added code performs the same test as UW imapd; it
1692 checks after acquiring the lock that its file descriptor still refers to
1693 the same named file.
1695 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1696 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1699 11. Added the smtp_active_hostname option. If used, this will typically be made
1700 to depend on the incoming interface address. Because $interface_address is
1701 not set up until the daemon has forked a reception process, error responses
1702 that can happen earlier (such as "too many connections") no longer contain
1705 12. If an expansion in a condition on a "warn" statement fails because a lookup
1706 defers, the "warn" statement is abandoned, and the next ACL statement is
1707 processed. Previously this caused the whole ACL to be aborted.
1709 13. Added the iplsearch lookup type.
1711 14. Added ident_timeout as a log selector.
1713 15. Added tls_certificate_verified as a log selector.
1715 16. Added a global option tls_require_ciphers (compare the smtp transport
1716 option of the same name). This controls incoming TLS connections.
1718 17. I finally figured out how to make tls_require_ciphers do a similar thing
1719 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1720 before starting the TLS session.
1722 18. Tabs are now shown as \t in -bP output.
1724 19. If the log selector return_path_on_delivery was set, Exim crashed when
1725 bouncing a message because it had too many Received: header lines.
1727 20. If two routers both had headers_remove settings, and the first one included
1728 a superfluous trailing colon, the final name in the first list and the
1729 first name in the second list were incorrectly joined into one item (with a
1730 colon in the middle).
1736 1. Added -C and -D options to the exinext utility, mainly to make it easier
1737 to include in the automated testing, but these could be helpful when
1738 multiple configurations are in use.
1740 2. The exinext utility was not formatting the output nicely when there was
1741 an alternate port involved in the retry record key, nor when there was a
1742 message id as well (for retries that were specific to a specific message
1743 and a specific host). It was also confused by IPv6 addresses, because of
1744 the additional colons they contain. I have fixed the IPv4 problem, and
1745 patched it up to do a reasonable job for IPv6.
1747 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1748 delivery, the log line now contains "pipelined" if PIPELINING was used.
1750 4. An SMTP transport process used to panic and die if the bind() call to set
1751 an explicit outgoing interface failed. This has been changed; it is now
1752 treated in the same way as a connect() failure.
1754 5. A reference to $sender_host_name in the part of a conditional expansion
1755 that was being skipped was still causing a DNS lookup. This no longer
1758 6. The def: expansion condition was not recognizing references to header lines
1759 that used bh_ and bheader_.
1761 7. Added the _cache feature to named lists.
1763 8. The code for checking quota_filecount in the appendfile transport was
1764 allowing one more file than it should have been.
1766 9. For compatibility with Sendmail, the command line option
1774 and sets the incoming protocol and host name (for trusted callers). The
1775 host name and its colon can be omitted when only the protocol is to be set.
1776 Note the Exim already has two private options, -pd and -ps, that refer to
1777 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1778 "s", but I don't think that's a major issue.
1780 10. A number of refactoring changes to the code, none of which should affect
1783 (a) The number of logging options was getting close to filling up the
1784 32-bit word that was used as a bit map. I have split them into two classes:
1785 those that are passed in the argument to log_write(), and those that are
1786 only ever tested independently outside of that function. These are now in
1787 separate 32-bit words, so there is plenty of room for expansion again.
1788 There is no change in the user interface or the logging behaviour.
1790 (b) When building, for example, log lines, the code previously used a
1791 macro that called string_cat() twice, in order to add two strings. This is
1792 not really sufficiently general. Furthermore, there was one instance where
1793 it was actually wrong because one of the argument was used twice, and in
1794 one call a function was used. (As it happened, calling the function twice
1795 did not affect the overall behaviour.) The macro has been replaced by a
1796 function that can join an arbitrary number of extra strings onto a growing
1799 (c) The code for expansion conditions now uses a table and a binary chop
1800 instead of a serial search (which was left over from when there were very
1801 few conditions). Also, it now recognizes conditions like "pam" even when
1802 the relevant support is not compiled in: a suitably worded error message is
1803 given if an attempt is made to use such a condition.
1805 11. Added ${time_interval:xxxxx}.
1807 12. A bug was causing one of the ddress fields not to be passed back correctly
1808 from remote delivery subprocesses. The field in question was not being
1809 subsequently used, so this caused to problems in practice.
1811 13. Added new log selectors queue_time and deliver_time.
1813 14. Might have fixed a bug in maildirsizefile handling that threw up
1814 "unexpected character" debug warnings, and recalculated the data
1815 unnecessarily. In any case, I expanded the warning message to give more
1818 15. Added the message "Restricted characters in address" to the statements in
1819 the default ACL that block characters like @ and % in local parts.
1821 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1822 Three changes have been made:
1824 (a) There was a serious bug; a negative response to MAIL caused the whole
1825 recipient domain to be cached as invalid, thereby blocking all messages
1826 to all local parts at the same domain, from all senders. This bug has
1827 been fixed. The domain is no longer cached after a negative response to
1828 MAIL if the sender used is not empty.
1830 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1833 (c) A new callout option, "use_sender" has been added for people who want
1834 the modified behaviour.
1840 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1843 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1844 indeed breaks things for older releases.
1846 3. Added additional logging to the case where there is a problem reading data
1847 from a filter that is running in a subprocess using a pipe, in order to
1848 try to track down a specific problem.
1850 4. Testing facility fudge: when running in the test harness and attempting
1851 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1852 getting "No route to host". Convert this to a timeout.
1854 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1857 6. Some OS don't have socklen_t but use size_t instead. This affects the
1858 fifth argument of getsockopt() amongst other things. This is now
1859 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1860 can be set for individual OS. I have set it for SunOS5, OSF1, and
1861 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1862 some earlier ones do not.
1864 7. Change 4.30/15 was not doing the test caselessly.
1866 8. The standard form for an IPv6 address literal was being rejected by address
1867 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1868 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1869 this, as well as the form without the "IPv6" on the front (but only when
1870 address literals are enabled, of course).
1872 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1874 10. Exim crashed if a message with an empty sender address specified by -f
1875 encountered a router with an errors_to setting. This could be provoked only
1876 by a command such as
1880 where an empty string was supplied; "<>" did not hit this bug.
1882 11. Installed PCRE release 4.5.
1884 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1885 remained set. It is now erased.
1887 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1888 times from message ids (which are base 36 rather than the normal 62).
1890 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1891 were being counted as actual protocol errors, and logged if the log
1892 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1893 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1894 and DATA following a set of rejected RCPTs do not count as protocol errors.
1895 In other words, Exim assumes they were pipelined, though this may not
1896 actually be the case. Of course, in all cases the client gets an
1897 appropriate error code.
1899 15. If a lookup fails in an ACL condition, a message about the failure may
1900 be available; it is used if testing the ACL cannot continue, because most
1901 such messages specify what the cause of the deferral is. However, some
1902 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1903 that caused an old message to be retained and used if a later statement
1904 caused a defer, replacing the real cause of the deferral.
1906 16. If an IP address had so many PTR records that the DNS lookup buffer
1907 was not large enough to hold them, Exim could crash while trying to process
1908 the truncated data. It now detects and logs this case.
1910 17. Further to 4.21/58, another change has been made: if (and only if) the
1911 first line of a message (the first header line) ends with CRLF, a bare LF
1912 in a subsequent header line has a space inserted after it, so as not to
1913 terminate the header.
1915 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1916 unnecessarily, used atoi() instead of strtol(), and didn't check the
1917 termination when getting file sizes from file names by regex.
1919 19. Completely re-implemented the support for maildirsize files, in the light
1920 of a number of problems with the previous contributed implementation
1921 (4.30/29). In particular:
1923 . If the quota is zero, the maildirsize file is maintained, but no quota is
1926 . If the maildir directory does not exist, it is created before any attempt
1927 to write a maildirsize file.
1929 . The quota value in the file is just a cache; if the quota is changed in
1930 the transport, the new value overrides.
1932 . A regular expression is available for excluding directories from the
1935 20. The autoreply transport checks the characters in options that define the
1936 message's headers; it allows continued headers, but it was checking with
1937 isspace() after an embedded newline instead of explicitly looking for a
1940 21. If all the "regular" hosts to which an address was routed had passed their
1941 expiry times, and had not reached their retry times, the address was
1942 bounced, even if fallback hosts were defined. Now Exim should go on to try
1945 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1946 equivalent code in the SMTP transport. Some hosts send humungous responses
1947 to HELO/EHLO, more than 1024 it seems.
1949 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1950 gives compiler warnings in some environments. I've now done it "properly",
1953 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1954 (because of problems with the built-in one) was declared to return uschar *
1955 instead of char *, causing compiler failure.
1957 25. Fixed a file descriptor leak when processing alias/forward files.
1959 26. Fixed a minor format string issue in dbfn.c.
1961 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1963 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1964 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1967 29. When a sender address is verified, it is cached, to save repeating the test
1968 when there is more than one recipient in a message. However, when the
1969 verification involves a callout, it is possible for different callout
1970 options to be set for different recipients. It is too complicated to keep
1971 track of this in the cache, so now Exim always runs a verification when a
1972 callout is required, relying on the callout cache for the optimization.
1973 The overhead is duplication of the address routing, but this should not be
1976 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
1977 to be verified with callout=postmaster, and the main callout worked but the
1978 postmaster check failed, the verification correctly failed. However, if a
1979 subsequent RCPT command asked for sender verification *without* the
1980 postmaster check, incorrect caching caused this verification also to fail,
1983 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1984 it was not caching the DNS options (qualify_single, search_parents) that
1985 were used when the lookup failed. A subsequent lookup with different
1986 options therefore always gave the same answer, though there were cases
1987 where it should not have. (Example: a "domains = !$mx_any" option on a
1988 dnslookup router: the "domains" option is always processed without any
1989 widening, but the router might have qualify_single set.) Now Exim uses the
1990 cached value only when the same options are set.
1992 32. Added John Jetmore's "exipick" utility to the distribution.
1994 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
1995 than a timeout (e.g. a certificate is required, and is not provided), an
1996 Exim server now closes the connection immediately. Previously it waited for
1997 the client to close - but if the client is SSL, it seems that they each
1998 wait for each other, leading to a delay before one of them times out.
2000 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2001 maintained 0.8.x compatibility because I don't think many are using it, and
2002 it is clearly obsolete.
2004 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2007 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2008 client certificate was expired. A simple patch fixes this, though I don't
2009 understand the full logic of why the verify callback is called multiple
2012 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2013 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2014 which causes problems with some clients (such as the Certicom SSL Plus
2015 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2016 disables the coutermeasure allowing Eudora to connect."
2018 38. Exim was not checking that a write() to a log file succeeded. This could
2019 lead to Bad Things if a log got too big, in particular if it hit a file
2020 size limit. Exim now panics and dies if it cannot write to a log file, just
2021 as it does if it cannot open a log file.
2023 39. Modified OS/Makefile-Linux so that it now contains
2025 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2027 The two -D definitions ensure that Exim is compiled with large file
2028 support, which makes it possible to handle log files that are bigger than
2031 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2032 instance) a domain was checked against a named list that involved a lookup,
2033 causing $domain_data to be set, then another domain was checked against the
2034 same list, then the first domain was re-checked, the value of $domain_data
2035 after the final check could be wrong. In particular, if the second check
2036 failed, it could be set empty. This bug probably also applied to
2039 41. The strip_trailing_dot option was not being applied to the address given
2040 with the -f command-line option.
2042 42. The code for reading a message's header from the spool was incrementing
2043 $received_count, but never initializing it. This meant that the value was
2044 incorrect (doubled) while delivering a message in the same process in which
2045 it was received. In the most common configuration of Exim, this never
2046 happens - a fresh exec is done - but it can happen when
2047 deliver_drop_privilege is set.
2049 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2050 it now includes up to 150 characters of the unexpected data in the log
2053 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2054 and building data strings. The size of both of these buffers was 10 000
2055 bytes - far larger than anybody would *ever* want, thought I. Needless to
2056 say, somebody hit the limit. I have increased the maximum line length to
2057 20 000 and the maximum data length of concatenated lines to 100 000. I have
2058 also fixed two bugs, because there was no checking on these buffers. Tsk,
2059 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2060 buffer is too small.
2062 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2063 lsearch lookups. Now it does.
2065 46. When parsing a route_list item in a manualroute router, a fixed-length
2066 buffer was used for the list of hosts. I made this 1024 bytes long,
2067 thinking that nobody would ever have a list of hosts that long. Wrong.
2068 Somebody had a whole pile of complicated expansion conditions, and the
2069 string was silently truncated, leading to an expansion error. It turns out
2070 that it is easier to change to an unlimited length (owing to other changes
2071 that have happened since this code was originally written) than to build
2072 structure for giving a limitation error. The length of the item that
2073 expands into the list of hosts is now unlimited.
2075 47. The lsearch lookup could not handle data where the length of text line was
2076 more than 4095 characters. Such lines were truncated, leading to shortened
2077 data being returned. It should now handle lines of any length.
2079 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2080 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2083 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2084 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2085 no longer happens in comment lines. A list of replacements is now placed
2086 at the head of all of the source files, except those whose only change is
2087 to replace PERL_COMMAND in the very first #! line.
2089 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2090 messages on the queue, with a bottom-up merge sort, using code contributed
2091 by Michael Haardt. This should make operations like -bp somewhat faster on
2092 large queues. It won't affect queue runners, except when queue_run_in_order
2095 51. Installed eximstats 1.31 in the distribution.
2097 52. Added support for SRV lookups to the dnslookup router.
2099 53. If an ACL referred to $message_body or $message_body_end, the value was not
2100 reset for any messages that followed in the same SMTP session.
2102 54. The store-handling optimization for building very long strings was not
2103 differentiating between the different store pools. I don't think this
2104 actually made any difference in practice, but I've tidied it.
2106 55. While running the routers to verify a sender address, $sender_address
2107 was still set to the sender address. This is wrong, because when routing to
2108 send a bounce to the sender, it would be empty. Therefore, I have changed
2109 it so that, while verifying a sender address, $sender_address is set to <>.
2110 (There is no change to what happens when verifying a recipient address.)
2112 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2113 target A or AAAA records (if not already returned) without resetting the
2114 qualify_single or search_parents options of the DNS resolver. These are
2115 inappropriate in this case because the targets of MX and SRV records must
2116 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2117 target that, when qualified, matched something in the local domain. These
2118 two options are now turned off when doing these lookups.
2120 57. It seems that at least some releases of Reiserfs (which does not have the
2121 concept of a fixed number of inodes) returns zero and not -1 for the
2122 number of available inodes. This interacted badly with check_spool_inodes,
2123 which assumed that -1 was the "no such thing" setting. What I have done is
2124 to check that the total number of inodes is greater than zero before doing
2125 the test of how many are available.
2127 58. When a "warn" ACL statement has a log_message modifier, the message is
2128 remembered, and not repeated. This is to avoid a lot of repetition when a
2129 message has many recipients that cause the same warning to be written.
2130 Howewer, Exim was preserving the list of already written lines for an
2131 entire SMTP session, which doesn't seem right. The memory is now reset if a
2132 new message is started.
2134 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2135 debugging output unless log_rewrite was also set.
2137 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2138 of (int)(handle) when we know that handle contains (void *)(-1).
2140 61. The Exim daemon panic-logs an error return when it closes the incoming
2141 connection. However "connection reset by peer" seems to be common, and
2142 isn't really an error worthy of noting specially, so that particular error
2145 62. When Exim is trying to find all the local interfaces, it used to panic and
2146 die if the ioctl to get the interface flags failed. However, it seems that
2147 on at least one OS (Solaris 9) it is possible to have an interface that is
2148 included in the list of interfaces, but for which you get a failure error
2149 for this call. This happens when the interface is not "plumbed" into a
2150 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2151 failure of the "get flags" call assumes that the interface is down.
2153 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2154 makes life easier for people who are doing arithmetic on fields extracted
2155 from dates, where you often get leading zeros that should not be
2156 interpreted as octal.
2158 64. Added qualify_domain to the redirect router, to override the global
2161 65. If a pathologically long header line contained very many addresses (the
2162 report of this problem mentioned 10 000) and each of them was rewritten,
2163 Exim could use up a very large amount of memory. (It kept on making new
2164 copies of the header line as it rewrote, and never released the old ones.)
2165 At the expense of a bit more processing, the header rewriting function has
2166 been changed so that it no longer eats memory in this way.
2168 66. The generation of the Received: header has been moved from the time that a
2169 message starts to be received, to the time that it finishes. The timestamp
2170 in the Received: header should now be very close to that of the <= log
2171 line. There are two side-effects of this change:
2173 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2174 logged header lines no longer include the local Received: line, because
2175 it has not yet been created. The same applies to a copy of the message
2176 that is returned to a non-SMTP sender when a message is rejected.
2178 (b) When a filter file is tested using -bf, no additional Received: header
2179 is added to the test message. After some thought, I decided that this
2182 This change does not affect the value of $received_for. It is still set
2183 after address rewriting, but before local_scan() is called.
2185 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2187 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2188 gave an unhelpful panic error message, and a defer error. I have managed to
2189 change this behaviour so that it now rejects any supplied certificate,
2190 which seems right, as the list of acceptable certificates is empty.
2192 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2193 gave an unhelpful defer error. I have not managed to make this reject any
2194 supplied certificates, but the error message it gives is "no certificate
2195 supplied", which is not helpful.
2197 70. exigrep's output now also includes lines that are not associated with any
2198 message, but which match the given pattern. Implemented by a patch from
2199 Martin Sluka, which also tidied up the Perl a bit.
2201 71. Recipient callout verification, like sender verification, was using <> in
2202 the MAIL FROM command. This isn't really the right thing, since the actual
2203 sender may affect whether the remote host accepts the recipient or not. I
2204 have changed it to use the actual sender in the callout; this means that
2205 the cache record is now keyed on a recipient/sender pair, not just the
2206 recipient address. There doesn't seem to be a real danger of callout loops,
2207 since a callout by the remote host to check the sender would use <>.
2208 [SEE ABOVE: changed after hitting problems.]
2210 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2211 temporary errors. However, in the case of such a code being given after
2212 the end of a data transmission (i.e. after ".") Exim was failing to write
2213 a retry record for the message. (Yes, there was some broken host that was
2214 actually sending 8xx at this point.)
2216 73. An unknown lookup type in a host list could cause Exim to panic-die when
2217 the list was checked. (An example that provoked this was putting <; in the
2218 middle of a list instead of at the start.) If this happened during a DATA
2219 ACL check, a -D file could be left lying around. This kind of configuration
2220 error no longer causes Exim to die; instead it causes a defer errror. The
2221 incident is still logged to the main and panic logs.
2223 74. Buglet left over from Exim 3 conversion. The message "too many messages
2224 in one connection" was written to the rejectlog but not the mainlog, except
2225 when address rewriting (yes!) was being logged.
2227 75. Added write_rejectlog option.
2229 76. When a system filter was run not as root (that is, when system_filter_user
2230 was set), the values of the $n variables were not being returned to the
2231 main process; thus, they were not subsequently available in the $sn
2234 77. Added +return_path_on_delivery log selector.
2236 78. A connection timeout was being treated differently from recipients deferred
2237 when testing hosts_max_try with a message that was older than the host's
2238 retry timeout. (The host should not be counted, thus allowing all hosts to
2239 be tried at least once before bouncing.) This may have been the cause of an
2240 occasionally reported bug whereby a message would remain on the queue
2241 longer than the retry timeout, but would be bounced if a delivery was
2242 forced. I say "may" because I never totally pinned down the problem;
2243 setting up timeout/retry tests is difficult. See also the next item.
2245 79. The ultimate address timeout was not being applied to errors that involved
2246 a combination of host plus message (for example, a timeout on a MAIL
2247 command). When an address resolved to a number of possible hosts, and they
2248 were not all tried for each delivery (e.g. because of hosts_max_try), a
2249 message could remain on the queue longer than the retry timeout.
2251 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2254 81. Fixed an obscure SMTP outgoing bug which required at least the following
2255 conditions: (a) there was another message waiting for the same server;
2256 (b) the server returned 5xx to all RCPT commands in the first message so
2257 that the message was not completed; (c) the server dropped the connection
2258 or gave a negative response to the RSET that Exim sends to abort the
2259 transaction. The observed case was a dropped connection after DATA that had
2260 been sent in pipelining mode. That is, the server had advertised PIPELINING
2261 but was not implementing it correctly. The effect of the bug was incorrect
2262 behaviour, such as trying another host, and this could lead to a crash.
2268 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2269 and daemon.c were passed as pointers to ints; they should have been
2270 pointers to socklen_t variables (which are typically unsigned ints).
2272 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2275 3. Fixed a really odd bug that affected only the testing scheme; patching a
2276 certain fixed string in the binary changed the value of another string that
2277 happened to be identical to the end of the original first string.
2279 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2280 name", it returns that address as the IP address. On some operating
2281 systems (e.g. Solaris), it also passes back the IP address string as the
2282 "host name". However, on others (e.g. Linux), it passes back an empty
2283 string. Exim wasn't checking for this, and was changing the host name to an
2284 empty string, assuming it had been canonicized.
2286 5. Although rare, it is permitted to have more than one PTR record for a given
2287 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2288 all the names associated with an address, because they do in Solaris.
2289 However, it seems that they do not in Linux for data that comes from the
2290 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2291 I found this out when I moved to a new Linux workstation and tried to run
2292 the Exim test suite.
2294 To get round this problem I have changed the code so that it now does its
2295 own call to the DNS to look up PTR records when searching for a host name.
2296 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2297 addresses that are only in /etc/hosts are still found.
2299 This behaviour is, however, controlled by an option called host_lookup_
2300 order, which defaults to "bydns:byaddr". If people want to use the other
2301 order, or indeed, just use one or the other means of lookup, they can
2302 specify it in this variable.
2304 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2305 some operating systems, this comes back from gethostbyaddr() as an empty
2306 string, and this is what Exim used to test for. However, it seems that in
2307 other systems, "." is yielded. Exim now tests for this case too.
2309 7. The values of check_spool_space and check_log_space are now held internally
2310 as a number of kilobytes instead of an absolute number of bytes. If a
2311 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2312 kilobyte. This means that much larger values can be stored.
2314 8. Exim monitor: an attempt to get the action menu when not actually pointing
2315 at a message produces an empty menu entitled "No message selected". This
2316 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2317 no entries in it ("Shell widget menu has zero width and/or height"). So I
2318 have added a single, blank menu entry in this case.
2320 9. Added ${quote_local_part.
2322 10. MIME decoding is now applied to the contents of Subject: header lines when
2325 11. Now that a reference to $sender_host_address automatically causes a reverse
2326 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2327 host lookup before query-style lookups in lists that might use this
2328 variable. This has therefore been abolished, and the "net-" prefix is no
2329 longer necessary for query-style lookups.
2331 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2332 have been a typo for LFLAGS, so it has been changed.
2334 13. The install script calls Exim with "-C /dev/null" in order to find the
2335 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2336 to be output. Howeve, since Exim outputs its version number before the
2337 error, it didn't break the script. It just looked ugly. I fixed this by
2338 always allowing "-C /dev/null" if the caller is root.
2340 14. Ignore overlarge ACL variable number when reading spool file - insurance
2341 against a later release with more variables having written the file.
2343 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2344 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2345 this, as well as the form without the "IPv6" on the front.
2347 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2348 OS/Makefile-Darwin file.
2350 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2351 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2353 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2354 variable. However, some OS use ints for inodes. I've added cast to long int
2355 to get rid of the compiler warning.
2357 19. I had forgotten to lock out "/../" in configuration file names when
2358 ALT_CONFIG_PREFIX was set.
2360 20. Routers used for verification do not need to specify transports. However,
2361 if such a router generated a host list, and callout was configured, Exim
2362 crashed, because it could not find a port number from the (non-existent)
2363 transport. It now assumes port 25 in this circumstance.
2365 21. Added the -t option to exigrep.
2367 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2368 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2369 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2370 them is set, LOOKUP_LSEARCH is forced.
2372 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2374 24. Added sender and host information to the "rejected by local_scan()" log
2375 line; previously there was no indication of these.
2377 25. Added .include_if_exists.
2379 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2380 Linux. It turns out that not all file systems support this. Apparently some
2381 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2382 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2383 not supported on the file descriptor, is now ignored when Exim is trying to
2384 sync a directory. This applies only to Linux.
2386 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2388 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2389 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2392 29. Added support for maildirsize files from supplied patch (modified a bit).
2394 30. The use of :fail: followed by an empty string could lead Exim to respond to
2395 sender verification failures with (e.g.):
2397 550 Verification failed for <xxx>
2398 550 Sender verify failed
2400 where the first response line was missing the '-' that indicates it is not
2401 the final line of the response.
2403 31. The loop for finding the name of the user that called Exim had a hardwired
2404 limit of 10; it now uses the value of finduser_retries, which is used for
2405 all other user lookups.
2407 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2410 33. Exim was neglecting to zero errno before one call of strtol() when
2411 expanding a string and expecting an integer value. On some systems this
2412 resulted in spurious "integer overflow" errors. Also, it was casting the
2413 result into an int without checking.
2415 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2416 did not work. The code looks as if it has *never* worked, though it appears
2417 to have been documented since at least releast 1.62. I have made it work.
2419 35. The "timeout_DNS" error in retry rules, also documented since at least
2420 1.62, also never worked. As it isn't clear exactly what this means, and
2421 clearly it isn't a major issue, I have abolished the feature by treating it
2422 as "timeout", and writing a warning to the main and panic logs.
2424 36. The display of retry rules for -brt wasn't always showing the error code
2427 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2428 timeout_connect_A, timeout_connect_MX.
2430 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2431 to the empty sender.
2433 39. The daemon was not analysing the content of -oX till after it had closed
2434 stderr and disconnected from the controlling terminal. This meant that any
2435 syntax errors were only noted on the panic log, and the return code from
2436 the command was 0. By re-arranging the code a little, I've made the
2437 decoding happen first, so such errors now appear on stderr, and the return
2438 code is 1. However, the actual setting up of the sockets still happens in
2439 the disconnected process, so errors there are still only recorded on the
2442 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2443 connections (as happens on some IP stacks) was logged at start up time as
2444 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2445 it from "IPv6 and IPv4", which means that two separate sockets are being
2448 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2449 says whether AF_INET or AF_INET6 was passed as an argument.
2451 42. Exiwhat output was messed up when time zones were included in log
2454 43. Exiwhat now gives more information about the daemon's listening ports,
2455 and whether -tls-on-connect was used.
2457 44. The "port" option of the smtp transport is now expanded.
2459 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2460 silently ignored. Now an error message is written to the main and panic
2463 46. There's a new ACL modifier called "logwrite" which writes to a log file
2464 as soon as it is encountered.
2466 47. Added $local_user_uid and $local_user_gid at routing time.
2468 48. Exim crashed when trying to verify a sender address that was being
2471 49. Exim was recognizing only a space character after ".include". It now also
2472 recognizes a tab character.
2474 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2475 extracting the relevant information from the specification. The man page no
2476 longer contains scrambled data for the -d option, and I've added a section
2477 at the front about calling Exim under different names.
2479 51. Added "extra_headers" argument to the "mail" command in filter files.
2481 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2484 53. Installed eximstats 1.29.
2486 54. Added transport_filter_timeout as a generic transport option.
2488 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2489 Cc: header lines. This was required by RFC 822, but it not required by RFC
2492 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2493 incoming messages that did not have them. Now it does so only if the
2494 message originates locally, that is, if there is no associated remote host
2495 address. When Resent- header lines are present, this applies to the Resent-
2496 lines rather than the non-Resent- lines.
2498 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2499 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2501 58. Messages for configuration errors now include the name of the main
2502 configuration file - useful now that there may be more than one file in a
2503 list (.included file names were always shown).
2505 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2506 for those rare installations that do not start the daemon as root or run it
2507 setuid root. I've cut out the call to initgroups() if the daemon is not
2510 60. The Exim user and group can now be bound into the binary as text strings
2511 that are looked up at the start of Exim's processing.
2513 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2515 62. Added $mailstore_basename variable.
2517 63. Installed patch to sieve.c from Michael Haardt.
2519 64. When Exim failed to open the panic log after failing to open the main log,
2520 the original message it was trying to log was written to stderr and debug
2521 output, but if they were not available (the usual case in production), it
2522 was lost. Now it is written to syslog before the two lines that record the
2523 failures to open the logs.
2525 65. Users' Exim filters run in subprocesses under the user's uid. It is
2526 possible for a "deliver" command or an alias in a "personal" command to
2527 provoke an address rewrite. If logging of address rewriting is configured,
2528 this fails because the process is not running as root or exim. There may be
2529 a better way of dealing with this, but for the moment (because 4.30 needs
2530 to be released), I have disabled address rewrite logging when running a
2531 filter in a non-root, non-exim process.
2537 1. The buildconfig auxiliary program wasn't quoting the value set for
2538 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2539 not defined. This bug was masked in 4.22 by the effect that was fixed in
2542 2. Some messages that were rejected after a message id was allocated were
2543 shown as "incomplete" by exigrep. It no longer does this for messages that
2544 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2546 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2547 have allow_domain_literals set, the ID did not get logged in the <= line.
2548 Domain literals are now always recognized in Message-ID: header lines.
2550 4. The first argument for a ${extract expansion item is the key name or field
2551 number. Leading and trailing spaces in this item were not being ignored,
2552 causing some misleading effects.
2554 5. When deliver_drop_privilege was set, single queue runner processes started
2555 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2556 same command in the process it spins off) were not dropping privilege.
2558 6. When the daemon running as "exim" started a queue runner, it always
2559 re-executed Exim in the spun-off process. This is a waste of effort when
2560 deliver_drop_privilege is set. The new process now just calls the
2561 queue-runner function directly.
2567 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2570 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2571 ignored. Though the use of -oP was forcing the writing of a pid file, it
2572 was always written to the default place.
2574 3. If the message "no IP address found for host xxxx" is generated during
2575 incoming verification, it is now followed by identification of the incoming
2576 connection (so you can more easily find what provoked it).
2578 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2580 5. Added some features to "harden" Exim a bit more against certain attacks:
2582 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2583 be put in Local/Makefile. This is like the never_users runtime option,
2584 but it cannot be overridden. The default setting is "root".
2586 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2587 prefix string with which any file named in a -C command line option
2590 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2591 is retained for -C and -D only if the caller of Exim is root. Without
2592 it, the exim user may also use -C and -D and retain privilege.
2594 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2595 command line option is disabled.
2597 6. Macro names set by the -D option must start with an upper case letter, just
2598 like macro names defined in the configuration file.
2600 7. Added "dereference=" facility to LDAP.
2602 8. Two instances of the typo "uknown" in the source files are fixed.
2604 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2605 the Configure-Makefile script screwed up while processing it.
2607 10. Incorporated PCRE 4.4.
2609 11. The SMTP synchronization check was not operating right at the start of an
2610 SMTP session. For example, it could not catch a HELO sent before the client
2611 waited for the greeting. There is now a check for outstanding input at the
2612 point when the greeting is written. Because of the duplex, asynchronous
2613 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2614 way, but not yet received, when the check is performed.
2616 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2617 on TCP/IP sockets, because this apparently causes some broken clients to
2620 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2621 unchanged) from the Cygwin maintainer.
2623 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2624 was supported instead of testing for mbx. Effectively a typo.
2626 15. The spa authenticator server code was not checking that the input it
2627 received was valid base64.
2629 16. The debug output line for the "set" modifier in ACLs was not showing the
2630 name of the variable that was being set.
2632 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2634 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2635 lookup on its own. Something else was needed to trigger the lookup. For
2636 example, a match in host_lookup or the need for a host name in a host list.
2637 Now, if $sender_host_name is referenced and the host name has not yet been
2638 looked up, a lookup is performed. If the lookup fails, the variable remains
2639 empty, and $host_lookup_failed is set to "1".
2641 19. Added "eqi" as a case-independent comparison operator.
2643 20. The saslauthd authentication condition could segfault if neither service
2644 nor realm was specified.
2646 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2647 error message that was logged was misleading, and incoming SMTP
2648 connections were dropped. The message is now more accurate, and temporary
2649 errors are given to SMTP connections.
2651 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2652 (except RSET) with a 421 error, until QUIT is received. However, it was
2653 failing to send a response to QUIT.
2655 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2656 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2657 ACL is now run after the helo_try_verify_hosts code.
2659 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2660 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2661 case-independent, so other case variants are also recognized). Apparently
2662 some systems use these upper case variants.
2664 25. If more than two messages were waiting for the same host, and a transport
2665 filter was specified for the transport, Exim sent two messages over the
2666 same TCP/IP connection, and then failed with "socket operation on non-
2667 socket" when it tried to send the third.
2669 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2671 27. The extern definition of crypt16() in expand.c was not being excluded when
2672 the OS had its own crypt16() function.
2674 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2675 as a preferred synonym for return_size_limit, both as an option and as an
2678 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2680 30. Changed the default configuration ACL to relax the local part checking rule
2681 for addresses that are not in any local domains. For these addresses,
2682 slashes and pipe symbols are allowed within local parts, but the sequence
2683 /../ is explicitly forbidden.
2685 31. SPA server authentication was not clearing the challenge buffer before
2688 32. log_message in a "warn" ACL statement was writing to the reject log as
2689 well as to the main log, which contradicts the documentation and doesn't
2690 seem right (because no rejection is happening). So I have stopped it.
2692 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2693 However, I am unable to do any testing of this.
2695 34. Fixed an infelicity in the appendfile transport. When checking directories
2696 for a mailbox, to see if any needed to be created, it was accidentally
2697 using path names with one or more superfluous leading slashes; tracing
2698 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2700 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2701 contents are added to the log line that is written for every discarded
2702 recipient. (Previously a log_message setting was ignored.)
2704 36. The ${quote: operator now quotes the string if it is empty.
2706 37. The install script runs exim in order to find its version number. If for
2707 some reason other than non-existence or emptiness, which it checks, it
2708 could not run './exim', it was installing it with an empty version number,
2709 i.e. as "exim-". This error state is now caught, and the installation is
2712 38. An argument was missing from the function that creates an error message
2713 when Exim fails to connect to the socket for saslauthd authentication.
2714 This could cause Exim to crash, or give a corrupted message.
2716 39. Added isip, isip4, and isip6 to ${if conditions.
2718 40. The ACL variables $acl_xx are now saved with the message, and can be
2719 accessed later in routers, transports, and filters.
2721 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2722 strings in the file are not string-expanded.
2724 42. If a MAIL command specified a SIZE value that was too large to fit into an
2725 int variable, the check against message_size_limit failed. Such values are
2726 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2727 day this will have to be increased, but I don't think I want to be around
2728 when emails are that large.
2735 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2736 iconv() is not standard in FreeBSD.
2738 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2739 IPv6 enabled. The observed symptom was a segmentation fault on return from
2740 the function os_common_find_running_interfaces() in src/os.c.
2742 3. In the check_special_case() function in daemon.c I had used "errno" as an
2743 argument name, which causes warnings on some systems. This was basically a
2744 typo, since it was named "eno" in the comments!
2746 4. The code that waits for the clock to tick (at a resolution of some fraction
2747 of a second) so as to ensure message-id uniqueness was always waiting for
2748 at least one whole tick, when it could have waited for less. [This is
2749 almost certainly not relevant at current processor speeds, where it is
2750 unlikely to ever wait at all. But we try to future-proof.]
2752 5. The function that sleeps for a time interval that includes fractions of a
2753 second contained a race. It did not block SIGALRM between setting the
2754 timer, and suspending (a couple of lines later). If the interval was short
2755 and the sigsuspend() was delayed until after it had expired, the suspension
2756 never ended. On busy systems this could lead to processes getting stuck for
2759 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2760 process, before it forks any delivery processes. The open lookup caching
2761 mechanism meant that the open file or database connection was passed into
2762 the delivery process. The problem was that delivery processes always tidy
2763 up cached lookup data. This could cause a problem for the next delivery
2764 process started by the queue runner, because the external queue runner
2765 process does not know about the closure. So the next delivery process
2766 still has data in the lookup cache. In the case of a file lookup, there was
2767 no problem because closing a file descriptor in a subprocess doesn't affect
2768 the parent. However, if the lookup was caching a connection to a database,
2769 the connection was closed, and the second delivery process was likely to
2770 see errors such as "PGSQL: query failed: server closed the connection
2771 unexpectedly". The problem has been fixed by closing all cached lookups
2772 in a queue runner before running a delivery process.
2774 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2775 seem to have the "const" qualifier which it has on other OS. I've
2778 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2779 *of the same type* (client or server) with the same public name that an
2780 error should be diagnosed.
2782 9. When Exim looked up a host name for an IP address, but failed to find the
2783 original IP address when looking up the host name (a safety check), it
2784 output the message "<ip address> does not match any IP for NULL", which was
2785 confusing, to say the least. The bug was that the host name should have
2786 appeared instead of "NULL".
2788 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2789 user that is built into the binary, and the -C or -D options is used, root
2790 privilege is dropped before the configuration file is read. In addition,
2791 logging is switched to stderr instead of the normal log files. If the
2792 configuration then re-defines the Exim user, the unprivileged environment
2793 is probably not what is expected, so Exim logs a panic warning message (but
2796 However, if deliver_drop_privilege is set, the unprivileged state may well
2797 be exactly what is intended, so the warning has been cut out in that case,
2798 and Exim is allowed to try to write to its normal log files.
2804 1. smtp_return_error_details was not giving details for temporary sender
2805 or receiver verification errors.
2807 2. Diagnose a configuration error if two authenticators have the same public
2810 3. Exim used not to create the message log file for a message until the first
2811 delivery attempt. This could be confusing when incoming messages were held
2812 for policy or load reasons. The message log file is now created at the time
2813 the message is received, and an initial "Received" line is written to it.
2815 4. The automatically generated man page for command line options had a minor
2816 bug that caused no ill effects; however, a more serious problem was that
2817 the procedure for building the man page automatically didn't always
2818 operate. Consequently, release 4.20 contains an out-of-date version. This
2819 shouldn't happen again.
2821 5. When building Exim with embedded Perl support, the script that builds the
2822 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2823 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2825 6. The freeze_tell option was not being used for messages that were frozen on
2826 arrival, either by an ACL or by local_scan().
2828 7. Added the smtp_incomplete_transaction log selector.
2830 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2831 was accepting AUTH without a new EHLO.
2833 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2834 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2837 10. Logging of TCP/IP connections (when configured) now happens in the main
2838 daemon process instead of the child process, so that the TCP/IP connection
2839 count is more accurate (but it can never be perfect).
2841 11. The use of "drop" in a nested ACL was not being handled correctly in the
2842 outer ACL. Now, if condition failure induced by the nested "drop" causes
2843 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2844 or "require"), the connection is dropped.
2846 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2847 that yield "discard" can now be used with an "accept" or a "discard" verb,
2848 but an error is generated for any others (because I can't see a useful way
2849 to define what should happen).
2851 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2852 are now processed in the same way as lines in the Exim configuration file.
2853 In particular, continuation lines are supported.
2855 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2857 15. Added -ti meaning -t -i.
2859 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2860 domains, and warn by logging if others are found.
2862 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2863 returned from GIFCONF when Exim is trying to find the list of interfaces on
2864 a host. The code in os.c has been modified to copy each ifreq to an aligned
2865 structure in all cases.
2867 Also, in some cases, the returned ifreq's were being copied to a 'struct
2868 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2869 means the last couple of bytes of an IPv6 address could be chopped if the
2870 ifreq contained only a normal sockaddr (14 bytes storage).
2872 18. Named domain lists were not supported in the hosts_treat_as_local option.
2873 An entry such as +xxxx was not recognized, and was treated as a literal
2876 19. Ensure that header lines added by a DATA ACL are included in the reject log
2877 if the ACL subsequently rejects the message.
2879 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2880 MD5 (which is deprecated).
2882 21. When testing a filter file using -bf, Exim was writing a message when it
2883 took the sender from a "From " line in the message, but it was not doing so
2884 when it took $return_path from a Return-Path: header line. It now does.
2886 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2887 with a valid header line field name (a series of printing characters
2888 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2890 23. Changed "disc" in the source to "disk" to conform to the documentation and
2891 the book and for uniformity.
2893 24. Ignore Sendmail's -Ooption=value command line item.
2895 25. When execve() failed while trying to run a command in a pipe transport,
2896 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2897 could be confused with a return value of 69 from the command itself. This
2898 has been changed to 127, the value the shell returns if it is asked to run
2899 a non-existent command. The wording for the related log line suggests a
2900 non-existent command as the problem.
2902 26. If received_header_text expands to an empty string, do not add a Received:
2903 header line to the message. (Well, it adds a token one on the spool, but
2904 marks it "old" so that it doesn't get used or transmitted.)
2906 27. Installed eximstats 1.28 (addition of -nt option).
2908 28. There was no check for failure on the call to getsockname() in the daemon
2909 code. This can fail if there is a shortage of resources on the system, with
2910 ENOMEM, for example. A temporary error is now given on failure.
2912 29. Contrary to the C standard, it seems that in some environments, the
2913 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2914 program. Exim now does this explicitly; it affects the formatting of
2915 timestamps using strftime().
2917 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2918 complaints. I've now initialized all the variables, to avoid this.
2920 32. Header lines added by a system filter were not being "seen" during
2921 transport-time rewrites.
2923 33. The info_callback() function passed to OpenSSL is set up with type void
2924 (*)(SSL *, int, int), as described somewhere. However, when calling the
2925 function (actually a macro) that sets it up, the type void(*)() is
2926 expected. I've put in a cast to prevent warnings from picky compilers.
2928 34. If a DNS black list lookup found a CNAME record, but there were no A
2929 records associated with the domain it pointed at, Exim crashed.
2931 35. If a DNS black list lookup returned more than one A record, Exim ignored
2932 all but the first. It now scans all returned addresses if a particular IP
2933 value is being sought. In this situation, the contents of the
2934 $dnslist_value variable are a list of all the addresses, separated by a
2937 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2938 to accept a host name and all its aliases if the forward lookup for any of
2939 them yielded the IP address of the incoming connection. Now it accepts only
2940 those names whose forward lookup yields the correct IP address. Any other
2941 names are discarded. This closes a loophole whereby a rogue DNS
2942 administrator could create reverse DNS records to break through a
2943 wildcarded host restriction in an ACL.
2945 37. If a user filter or a system filter that ran in a subprocess used any of
2946 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2947 the wrong values were passed to the pipe command ($thisaddress had the
2948 value of $0, $0 had the value of $1, etc). This bug was introduced by
2949 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2951 38. Improved the line breaking for long SMTP error messages from ACLs.
2952 Previously, if there was no break point between 40 and 75 characters, Exim
2953 left the rest of the message alone. Two changes have been made: (a) I've
2954 reduced the minimum length to 35 characters; (b) if it can't find a break
2955 point between 35 and 75 characters, it looks ahead and uses the first one
2956 that it finds. This may give the occasional overlong line, but at least the
2957 remaining text gets split now.
2959 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2960 file descriptors might be low, and that setting 1000 would always raise it.
2961 It turns out that in some environments, the limit is already over 1000 and
2962 that lowering it causes trouble. So now Exim takes care not to decrease it.
2964 40. When delivering a message, the value of $return_path is set to $sender_
2965 address at the start of routing (routers may change the value). By an
2966 oversight, this default was not being set up when an address was tested by
2967 -bt or -bv, which affected the outcome if any router or filter referred to
2970 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2971 log only when "message" or "log_message" are set. However, if one of the
2972 conditions was an address verification, or a call to a nested ACL, the
2973 messages generated by the underlying test were being passed through. This
2974 no longer happens. The underlying message is available in $acl_verify_
2975 message for both "message" and "log_message" expansions, so it can be
2976 passed through if needed.
2978 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2979 new expansion $bh_ to give the encoded byte string without charset
2980 translation. Translation happens only if iconv() is available; HAVE_ICONV
2981 indicates this at build time. HEADERS_CHARSET gives the charset to
2982 translate to; headers_charset can change it in the configuration, and
2983 "headers charset" can change it in an individual filter file.
2985 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
2986 that creates RFC 2047 encoded "words" labels them as that charset instead
2987 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2988 expansion operator; (ii) when Exim creates a From: line for a local
2989 message; (iii) when a header line is rewritten to include a "phrase" part.
2991 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2992 buggy, causing it to skip the first lines of messages whose message ID
2993 ended in 'D'. This would not have bitten before Exim release 4.14, because
2994 message IDs were unlikely to end in 'D' before then. The effect was to have
2995 incorrect size information for certain domains.
2997 45. #include "config.h" was missing at the start of the crypt16.c module. This
2998 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3001 46. If there was a timeout during a "random" callout check, Exim treated it as
3002 a failure of the random address, and carried on sending RSET and the real
3003 address. If the delay was just some slowness somewhere, the response to the
3004 original RCPT would be taken as a response to RSET and so on, causing
3005 mayhem of various kinds.
3007 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3008 when I implemented it. It didn't allow for the fact that some option values
3009 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3010 the right test for positive values.
3012 48. Domain names in DNS records are case-independent. Exim always looks them up
3013 in lower case. Some resolvers return domain names in exactly the case they
3014 appear in the zone file, that is, they may contain uppercase letters. Not
3015 all resolvers do this - some return always lower case. Exim was treating a
3016 change of case by a resolver as a change of domain, similar to a widening
3017 of a domain abbreviation. This triggered its re-routing code and so it was
3018 trying to route what was effectively the same domain again. This normally
3019 caused routing to fail (because the router wouldn't handle the domain
3020 twice). Now Exim checks for this case specially, and just changes the
3021 casing of the domain that it ultimately uses when it transmits the message
3024 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3027 50. If a filter generated a file delivery with a non-absolute name (possible if
3028 no home directory exists for the router), the forbid_file option was not
3031 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3032 the existing equality matching.
3034 52. Exim was using ints instead of ino_t variables in some places where it was
3035 dealing with inode numbers.
3037 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3038 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3039 called TMPDIR, and if it finds it is different, it changes its value.
3041 54. The smtp_printf() function is now made available to local_scan() so
3042 additional output lines can be written before returning. There is also an
3043 smtp_fflush() function to enable the detection of a dropped connection.
3044 The variables smtp_input and smtp_batched_input are exported to
3047 55. Changed the default runtime configuration: the message "Unknown user"
3048 has been removed from the ACL, and instead placed on the localuser router,
3049 using the cannot_route_message feature. This means that any verification
3050 failures that generate their own messages won't get overridden. Similarly,
3051 the "Unrouteable address" message that was in the ACL for unverifiable
3052 relay addresses has also been removed.
3054 56. Added hosts_avoid_esmtp to the smtp transport.
3056 57. The exicyclog script was not checking for the esoteric option
3057 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3058 will work only if exicyclog is run under the appropriate euid.
3060 58. Following a discussion on the list, the rules by which Exim recognises line
3061 endings on incoming messages have been changed. The -dropcr and drop_cr
3062 options are now no-ops, retained only for backwards compatibility. The
3063 following line terminators are recognized: LF CRLF CR. However, special
3064 processing applies to CR:
3066 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3067 nor a local message in the state where . is a terminator.
3069 (ii) If a bare CR is encountered in a header line, an extra space is added
3070 after the line terminator so as not to end the header. The reasoning
3071 behind this is that bare CRs in header lines are most likely either
3072 to be mistakes, or people trying to play silly games.
3074 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3075 was being incorrectly given as 18 bytes larger than it should have been.
3076 This is a VOB (very old bug).
3078 60. This may never have affected anything current, but just in case it has:
3079 When the local host is found other than at the start of a list of hosts,
3080 the local host, those with the same MX, and any that follow, are discarded.
3081 When the list in question was part of a longer list of hosts, the following
3082 hosts (not currently being processed) were also being discarded. This no
3083 longer happens. I'm not sure if this situation could ever has previously
3086 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3089 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3090 header. This is something that is recommended in a new Internet Draft, and
3091 is something that is documented as being done by Sendmail. There are two
3092 possible values. For messages generated by the autoreply transport, Exim
3095 Auto-Submitted: auto-replied
3097 whereas for all other generated messages (e.g. bounces) it adds
3099 Auto-Submitted: auto-generated
3101 63. The "personal" condition in filters now includes a test for the
3102 Auto-Submitted: header. If it contains the string "auto-" the message it
3103 not considered personal.
3105 64. Added rcpt_include_affixes as a generic transport option.
3107 65. Added queue_only_override (default true).
3109 66. Added the syslog_duplication option.
3111 67. If what should have been the first header line of a message consisted of
3112 a space followed by a colon, Exim was mis-interpreting it as a header line.
3113 It isn't of course - it is syntactically invalid and should therefore be
3114 treated as the start of the message body. The misbehaviour could have
3115 caused a number of strange effects, including loss of data in subsequent
3116 header lines, and spool format errors.
3118 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3119 client host had authenticated. This control can now be exercised by an ACL
3120 for more flexibility.
3122 69. By default, callouts do not happen when testing with -bh. There is now a
3123 variant, -bhc, which does actually run the callout code, including
3124 consulting and updating the callout cache.
3126 70. Added support for saslauthd authentication, courtesy of Alexander
3129 71. If statvfs() failed on the spool or log directories while checking their
3130 size for availability, Exim confusingly gave the error "space shortage".
3131 Furthermore, in debugging mode it crashed with a floating point exception.
3132 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3133 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3134 serious problem, Exim now writes to the main and panic logs when this
3135 happens, with details of the failure. It then refuses to accept the
3136 incoming message, giving the message "spool directory problem" or "log
3137 directory problem" with a 421 code for SMTP messages.
3139 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3140 0, 1, and 2 exist, because some OS complain for execs without them (see
3141 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3142 descriptors. However, the code omitted to check that the open succeeded,
3143 causing mysterious errors if for some reason the permissions on /dev/null
3144 got screwed. Now Exim writes a message to the main and panic logs, and
3145 bombs out if it can't open /dev/null.
3147 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3148 interact so that it is all more flexible. It is supposed to remain
3149 backwards compatible. Also added extra_local_interfaces.
3151 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3152 to bomb out with an assertion failure - to the client this appears as a
3153 connection drop. This problem occurs in the part of the code that was taken
3154 from the Samba project. Fortunately, the assertion is in a very simple
3155 function, so I have fixed this by reproducing the function inline in the
3156 one place where it is called, and arranging for authentication to fail
3157 instead of killing the process with assert().
3159 75. The SPA client code was not working when the server requested OEM rather
3160 than Unicode encoding.
3162 76. Added code to make require_files with a specific uid setting more usable in
3163 the case where statting the file as root fails - usually a non-root-mounted
3164 NFS file system. When this happens and the failure is EACCES, Exim now
3165 forks a subprocess and does the per-uid checking as the relevant uid.
3167 77. Added process_log_path.
3169 78. If log_file_path was not explicitly set, a setting of check_log_space or
3170 check_log_inodes was ignored.
3172 79. If a space check for the spool or log partitions fails, the incident is now
3173 logged. Of course, in the latter case the data may get lost...
3175 80. Added the %p formatting code to string_format() so that it can be used to
3176 print addresses in debug_print(). Adjusted all the address printing in the
3177 debugging in store.c to use %p rather than %d.
3179 81. There was a concern that a line of code in smtp_in.c could overflow a
3180 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3181 initially expressed, the concern was not well-founded, because trailing
3182 spaces are removed early. However, if the trailing spaces were followed by
3183 a NULL, they did not get removed, so the overflow was possible. Two fixes
3186 (a) I re-wrote the offending code in a cleaner fashion.
3187 (b) If an incoming SMTP command contains a NULL character, it is rejected
3190 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3191 runs initgroups(), so that if the Exim user is in any additional groups,
3192 they will be used during message reception.
3198 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob