1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.193 2005/08/02 11:22:23 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Added the "success_on_redirect" address verification option. See
10 NewStuff for rationale and an example.
12 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
14 PH/02 Patch to exigrep to allow it to work on syslog lines.
16 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
17 fread() to skip over the body file's header line, because in Cygwin the
18 header line is locked and is inaccessible.
20 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
21 co-exist for some time) to make it clear that it is the Exim ID that is
22 referenced, not the Message-ID: header line.
24 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
25 string_format() function, because snprintf() does not exist on all
28 PH/06 The use of forbid_filter_existstest now also locks out the use of the
29 ${stat: expansion item.
31 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
32 protocol synchronization error", to keep the pedants happy.
34 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
35 well as for IRIX systems, when gcc is being used. See the host.c source
38 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
40 PH/10 Named domain lists were not working if used in a queue_smtp_domains
43 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
44 transport and to the smtp transport in LMTP mode.
50 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
52 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
53 command itself ended in error, the underlying message about the transport
54 filter timeout was being overwritten with the pipe command error. Now the
55 underlying error message should be appended to the second error message.
57 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
59 PH/02 Reduce the amount of output that "make" produces by default. Full output
60 can still be requested.
62 PH/03 The warning log line about a condition test deferring for a "warn" verb
63 was being output only once per connection, rather than after each
64 occurrence (because it was using the same function as for successful
65 "warn" verbs). This seems wrong, so I have changed it.
67 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
68 it should not have, which might have caused a crash in the right
69 circumstances, but probably never did.
71 PH/04 Installed a modified version of Tony Finch's patch to make submission
72 mode fix the return path as well as the Sender: header line, and to
73 add a /name= option so that you can make the user's friendly name appear
76 TF/03 Added the control = fakedefer ACL modifier.
78 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
79 Mark Lowes for thorough testing.
81 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
83 TK/03 Merged latest SRS patch from Miles Wilton.
85 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
86 with the definition in sysexits.h (which is #included earlier).
87 Fortunately, Exim does not actually use EX_OK. The code used to try to
88 preserve the sysexits.h value, by assumimg that macro definitions were
89 scanned for macro replacements. I have been disabused of this notion,
90 so now the code just undefines EX_OK before #including unistd.h.
92 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
93 in the smtp transport. When a block could not be written in a single
94 write() function, the timeout was being re-applied to each part-write.
95 This seems wrong - if the receiver was accepting one byte at a time it
96 would take for ever. The timeout is now adjusted when this happens. It
97 doesn't have to be particularly precise.
99 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
100 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
102 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
103 without a domain if the check to <postmaster@domain> fails.
105 SC/01 Eximstats: added -xls and the ability to specify output files
106 (patch written by Frank Heydlauf).
108 SC/02 Eximstats: use FileHandles for outputing results.
110 SC/03 Eximstats: allow any combination of xls, txt, and html output.
112 SC/04 Eximstats: fixed display of large numbers with -nvr option
114 SC/05 Eximstats: fixed merging of reports with empty tables.
116 SC/06 Eximstats: added the -include_original_destination flag
118 SC/07 Eximstats: removed tabs and trailing whitespace.
120 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
122 TK/06 MBOX spool code: Add real "From " MBOX separator line
123 so the .eml file is really in mbox format (even though
124 most programs do not really care). Patch from Alex Miller.
126 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
127 The latter is generated from $received_to and is only set if the
128 message has one envelope recipient. SA can use these headers,
129 obviously out-of-the-box. Patch from Alex Miller.
131 PH/08 The ${def test on a variable was returning false if the variable's
132 value was "0", contrary to what the specification has always said!
133 The result should be true unless the variable is empty.
135 PH/09 The syntax error of a character other than { following "${if
136 def:variable_name" (after optional whitespace) was not being diagnosed.
137 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
138 accidental colon was present, for example, could give incorrect results.
140 PH/10 Tidied the code in a number of places where the st_size field of a stat()
141 result is used (not including appendfile, where other changes are about
144 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
145 This involved changing a lot of size variables from int to off_t. It
146 should work with maildirs and everything.
148 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
149 spamd dying while we are connected to it.
151 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
154 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
155 function for BDB 4.3.
157 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
159 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
160 the sources for PCRE have been re-arranged and more files are now
163 PH/15 The code I had for printing potentially long long variables in PH/11
164 above was not the best (it lost precision). The length of off_t variables
165 is now inspected at build time, and an appropriate printing format (%ld
166 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
167 to be "long long int" or "long int". This is needed for the internal
168 formatting function string_vformat().
170 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
171 the configuration file to be ":syslog", then the script "guesses" where
172 the logs files are, rather than using the compiled in default. In our
173 case the guess is not the same as the compiled default, so the script
174 suddenly stopped working when I started to use syslog. The patch checks
175 to see if log_file_path is "". If so, it attempts to read it from exim
176 with no configuration file to get the compiled in version, before it
177 falls back to the previous guessing code."
179 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
180 implementing BATV in an Exim configuration. See NewStuff for the gory
183 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
184 Makefile that are specific to HP-UX.
186 PH/18 If the "use_postmaster" option was set for a recipient callout together
187 with the "random" option, the postmaster address was used as the MAIL
188 FROM address for the random test, but not for the subsequent recipient
189 test. It is now used for both.
191 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
192 patch removes a few documentation additions to RFC 3028, because the
193 latest draft now contains them. It adds the new en;ascii-case comparator
194 and a new error check for 8bit text in MIME parts. Comparator and
195 require names are now matched exactly. I enabled the subaddress
196 extension, but it is not well tested yet (read: it works for me)."
198 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
199 rework some of the code of TK/09 above to avoid the hardwired use of
200 "%lld" and "long long". Replaced the call to snprintf() with a call to
203 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
204 records point to non-existent hosts", "retry timeout exceeded", and
205 "retry time not reached for any host after a long failure period".
207 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
208 experimental DomainKeys support:
210 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
211 (2) On an error such as an illegally used "control", the wrong name for
212 the control was given.
214 These problems did NOT occur unless DomainKeys support was compiled.
216 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
218 PH/24 Added ${if match_ip condition.
220 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
221 databases so that it will be absolutely obvious if a crash occurs in the
222 DB library. This is a regular occurrence (often caused by mis-matched
225 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
226 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
227 picked up on a user's system that detects such things. There doesn't seem
228 to be a gcc warning option for this - only an attribute that has to be
229 put on the function's prototype. It seems that in Fedora Core 4 they have
230 set this on a number of new functions. No doubt there will be more in due
233 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
234 specify a transport. However, if an address that was verified by such a
235 router was the subject of a callout, Exim crashed because it tried to
236 read the rcpt_include_affixes from the non-existent transport. Now it
237 just assumes that the setting of that option is false. This bug was
238 introduced by 4.51/PH/31.
240 PH/28 Changed -d+all to exclude +memory, because that information is very
241 rarely of interest, but it makes the output a lot bigger. People tend to
242 do -d+all out of habit.
244 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
245 code in os-type was giving problems when libc.so lives in lib64, like on
248 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
249 aren't the modern standard, and it seems that some systems' include files
250 don't always have them. Exim was already checking for some of the newer
251 ones like T_AAAA, and defining it itself. I've added checks for all the
252 record types that Exim uses.
254 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
255 not automatically generating a new one, as it is supposed to. This
256 prevented TLS from working. If the file did exist, but contained invalid
257 data, a new version was generated, as expected. It was only the case of a
258 non-existent file that was broken.
260 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
261 with a change in libdomainkeys > 0.64.
263 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
264 from DNS. If the selector record carries the flag, it now has
265 precedence over the domain-wide flag.
267 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
269 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
270 the use of a port name that isn't defined in /etc/services) occurred, the
271 message was deferred as in a normal delivery, and thus remained on the
272 spool, instead of being failed because of the mua_wrapper setting. This
273 is now fixed, and I tidied up some of the mua_wrapper messages at the
276 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
277 the messages in a hash of arrays rather than using individual hashes.
278 This is a bit cleaner and results in dramatic memory savings, albeit
279 at a slight CPU cost.
281 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
282 as requested by Marc Sherman.
284 SC/10 Eximstats: added histograms for user specified patterns as requested
287 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
289 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
290 fopen() in the content-scanning modules that did not already have it.
296 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
297 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
299 TK/02 Fix ACL "control" statement not being available in MIME ACL.
301 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
303 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
304 to test Sieve filters that use "vacation".
306 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
307 that changes the way the GnuTLS parameters are stored in the cache file.
308 The new format can be generated externally. For backward compatibility,
309 if the data in the cache doesn't make sense, Exim assumes it has read an
310 old-format file, and it generates new data and writes a new file. This
311 means that you can't go back to an older release without removing the
314 PH/03 A redirect router that has both "unseen" and "one_time" set does not
315 work if there are any delivery delays because "one_time" forces the
316 parent to be marked "delivered", so its unseen clone is never tried
317 again. For this reason, Exim now forbids the simultaneous setting of
320 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
321 redirected to themselves ("homonym" addresses). Read the long ChangeLog
322 entry if you want to know the details. The fix, however, neglected to
323 consider the case when local delivery batching is involved. The test for
324 "previously delivered" was not happening when checking to see if an
325 address could be batched with a previous (undelivered) one; under
326 certain circumstances this could lead to multiple deliveries to the same
329 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
330 in its include files, and this causes problems building Exim.
332 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
333 header_syntax) but Exim was just ignoring anything given after a slash.
334 In particular, this caused confusion with an attempt to use "verify =
335 reverse_host_lookup/defer_ok". An error is now given when options are
336 supplied for verify items that do not have them. (Maybe reverse_host_
337 lookup should have a defer_ok option, but that's a different point.)
339 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
340 defined by RFC 821) to 2048, because there were problems with some AUTH
341 commands, and RFC 1869 says the size should be increased for extended
342 SMTP commands that take arguments.
344 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
347 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
348 "unknown" error; now it says that the functionality isn't in the binary.
350 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
351 an address' error message when a string expansion fails (syntax or
352 whatever). Otherwise the password may appear in the log. Following change
353 PH/42 below, there is no longer a chance of it appearing in a bounce
356 PH/11 Installed exipick version 20050225.0 from John Jetmore.
358 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
359 of its addresses was ever tried. (Bugzilla bug #2.)
361 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
362 the result incorrectly in the debug output. (It correctly added a newline
363 to what was transported.)
365 TF/01 Added $received_time.
367 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
368 commented out examples of how to interface to a virus scanner and to
369 SpamAssassin. Also added commented examples of av_scanner and
370 spamd_address settings.
372 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
373 and controls are allowed in which ACLs. There were a couple of minor
374 errors. Some of the entries in the conditions table (which is a table of
375 where they are NOT allowed) were getting very unwieldy; rewrote them as a
376 negation of where the condition IS allowed.
378 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
380 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
381 header file does not have a version number, so I've had to invent a new
382 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
383 API. The code is untested by me (my Linux distribution still has 0.3.2 of
384 radiusclient), but it was contributed by a Radius user.
386 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
387 files or directories, for OpenSSL.
389 PH/19 When an Exim process that is running as root has to create an Exim log
390 file, it does so in a subprocess that runs as exim:exim so as to get the
391 ownership right at creation (otherwise, other Exim processes might see
392 the file with the wrong ownership). There was no test for failure of this
393 fork() call, which would lead to the process getting stuck as it waited
394 for a non-existent subprocess. Forks do occasionally fail when resources
395 run out. I reviewed all the other calls to fork(); they all seem to check
398 PH/20 When checking for unexpected SMTP input at connect time (before writing
399 the banner), Exim was not dealing correctly with a non-positive return
400 from the read() function. If the client had disconnected by this time,
401 the result was a log entry for a synchronization error with an empty
402 string after "input=" when read() returned zero. If read() returned -1
403 (an event I could not check), uninitialized data bytes were printed.
404 There were reports of junk text (parts of files, etc) appearing after
407 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
409 PH/22 Added support for macro redefinition, and (re)definition in between
410 driver and ACL definitions.
412 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
413 forgetting to use the resulting value; it was using the unexpanded value.
415 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
416 hadn't been configured. The fix is from Juergen Kreileder, who
417 understands it better than I do:
419 "Here's what I see happening with three configured cyrus_sasl
420 authenticators configured (plain, login, cram-md5):
422 On startup auth_cyrus_sasl_init() gets called for each of these.
423 This means three calls to sasl_listmech() without a specified mech_list.
424 => SASL tests which mechs of all available mechs actually work
425 => three warnings about OTP not working
426 => the returned list contains: plain, login, cram-md5, digest-md5, ...
428 With the patch, sasl_listmech() also gets called three times. But now
429 SASL's mech_list option is set to the server_mech specified in the the
430 authenticator. Or in other words, the answer from sasl_listmech()
431 gets limited to just the mech you're testing for (which is different
433 => the return list contains just 'plain' or 'login', 'cram-md5' or
434 nothing depending on the value of ob->server_mech.
436 I've just tested the patch: Authentication still works fine,
437 unavailable mechs specified in the exim configuration are still
438 caught, and the auth.log warnings about OTP are gone."
440 PH/25 When debugging is enabled, the contents of the command line are added
441 to the debugging output, even when log_selector=+arguments is not
444 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
445 answer is "GNU", and only if the return is "GNU/something" is the answer
448 PH/27 $acl_verify_message is now set immediately after the failure of a
449 verification in an ACL, and so is available in subsequent modifiers. In
450 particular, the message can be preserved by coding like this:
452 warn !verify = sender
453 set acl_m0 = $acl_verify_message
455 Previously, $acl_verify_message was set only while expanding "message"
456 and "log_message" when a very denied access.
458 PH/28 Modified OS/os.c-Linux with
460 -#ifndef OS_LOAD_AVERAGE
461 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
463 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
464 nomenclature these days.)
466 PH/29 Installed patch from the Sieve maintainer that adds the options
467 sieve_useraddress and sieve_subaddress to the redirect router.
469 PH/30 In these circumstances:
470 . Two addresses routed to the same list of hosts;
471 . First host does not offer TLS;
472 . First host accepts first address;
473 . First host gives temporary error to second address;
474 . Second host offers TLS and a TLS session is established;
475 . Second host accepts second address.
476 Exim incorrectly logged both deliveries with the TLS parameters (cipher
477 and peerdn, if requested) that were in fact used only for the second
480 PH/31 When doing a callout as part of verifying an address, Exim was not paying
481 attention to any local part prefix or suffix that was matched by the
482 router that accepted the address. It now behaves in the same way as it
483 does for delivery: the affixes are removed from the local part unless
484 rcpt_include_affixes is set on the transport.
486 PH/32 Add the sender address, as F=<...>, to the log line when logging a
487 timeout during the DATA phase of an incoming message.
489 PH/33 Sieve envelope tests were broken for match types other than :is. I have
490 applied a patch sanctioned by the Sieve maintainer.
492 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
493 the uid or gid is negative. A case of a negative gid caused this to be
494 noticed. The fix allows for either to be negative.
496 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
497 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
500 PH/36 The change PH/12 above was broken. Fixed it.
502 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
503 the grounds that routing the same address twice would always produce the
504 same answer. This might have been true once, but it is certainly no
505 longer true now. Routing a child address may depend on the previous
506 routing that produced that child. Some complicated redirection strategies
507 went wrong when messages had multiple recipients, and made Exim's
508 behaviour dependent on the order in which the addresses were given.
510 I have moved the duplicate checking until after the routing is complete.
511 Exim scans the addresses that are assigned to local and remote
512 transports, and removes any duplicates. This means that more work will be
513 done, as duplicates will always all be routed, but duplicates are
514 presumably rare, so I don't expect this is of any significance.
516 For deliveries to pipes, files, and autoreplies, the duplicate checking
517 still happens during the routing process, since they are not going to be
520 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
521 It corrects a timeout issue with spamd. This is Ian's comment: "The
522 background is that sometimes spamd either never reads data from a
523 connection it has accepted, or it never writes response data. The exiscan
524 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
525 blindly assumes that writes won't block so it may never time out."
527 PH/39 Allow G after quota size as well as K and M.
529 PH/40 The value set for $authenticated_id in an authenticator may not contain
530 binary zeroes or newlines because the value is written to log lines and
531 to spool files. There was no check on this. Now the value is run through
532 the string_printing() function so that such characters are converted to
533 printable escape sequences.
535 PH/41 $message_linecount is a new variable that contains the total number of
536 lines in the message. Compare $body_linecount, which is the count for the
539 PH/42 Exim no longer gives details of delivery errors for specific addresses in
540 bounce and delay warning messages, except in certain special cases, which
543 (a) An SMTP error message from a remote host;
544 (b) A message specified in a :fail: redirection;
545 (c) A message specified in a "fail" command in a system filter;
546 (d) A message specified in a FAIL return from the queryprogram router;
547 (e) A message specified by the cannot_route_message router option.
549 In these cases only, Exim does include the error details in bounce and
550 warning messages. There are also a few cases where bland messages such
551 as "unrouteable address" or "local delivery error" are given.
553 PH/43 $value is now also set for the "else" part of a ${run expansion.
555 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
556 being worked on, but at least Exim now implements the latest version to
559 PH/45 In a pipe transport, although a timeout while waiting for the pipe
560 process to complete was treated as a delivery failure, a timeout while
561 writing the message to the pipe was logged, but erroneously treated as a
562 successful delivery. Such timeouts include transport filter timeouts. For
563 consistency with the overall process timeout, these timeouts are now
564 treated as errors, giving rise to delivery failures by default. However,
565 there is now a new Boolean option for the pipe transport called
566 timeout_defer, which, if set TRUE, converts the failures into defers for
567 both kinds of timeout. A transport filter timeout is now identified in
570 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
571 systems where "make" and "gmake" are different, calling "gmake" at top
572 level broke things. I've arranged for the value of $(MAKE) to be passed
573 from the Makefile to this script so that it can call the same version of
577 A note about Exim versions 4.44 and 4.50
578 ----------------------------------------
580 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
581 changes of various kinds. As a consequence, a big documentation update was
582 needed. This delayed the release for rather longer than seemed good, especially
583 in the light of a couple of (minor) security issues. Therefore, the changes
584 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
585 release. So 4.44 and 4.50 are in effect two different branches that both start
588 I have left the 4.50 change log unchanged; it contains all the changes since
589 4.43. The change log for 4.44 is below; many of its items are identical to
590 those for 4.50. This seems to be the most sensible way to preserve the
591 historical information.
597 1. Minor wording change to the doc/README.SIEVE file.
599 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
600 computation of the current number of files was incorrect.
602 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
603 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
606 4. Give more explanation in the error message when the command for a transport
607 filter fails to execute.
609 5. There are several places where Exim runs a non-Exim command in a
610 subprocess. The SIGUSR1 signal should be disabled for these processes. This
611 was being done only for the command run by the queryprogram router. It is
612 now done for all such subprocesses. The other cases are: ${run, transport
613 filters, and the commands run by the lmtp and pipe transports.
615 6. Added CONFIGURE_GROUP build-time option.
617 7. Some older OS have a limit of 256 on the maximum number of file
618 descriptors. Exim was using setrlimit() to set 1000 as a large value
619 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
620 systems. I've change it so that if it can't get 1000, it tries for 256.
622 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
623 was an oversight, and furthermore, ever since the addition of extra
624 controls (e.g. 4.43/32), the checks on when to allow different forms of
625 "control" were broken. There should now be diagnostics for all cases when a
626 control that does not make sense is encountered.
628 9. Added the /retain_sender option to "control=submission".
630 10. $recipients is now available in the predata ACL (oversight).
632 11. Tidy the search cache before the fork to do a delivery from a message
633 received from the command line. Otherwise the child will trigger a lookup
634 failure and thereby defer the delivery if it tries to use (for example) a
635 cached ldap connection that the parent has called unbind on.
637 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
638 of $address_data from the recipient verification was clobbered by the
641 13. The value of address_data from a sender verification is now available in
642 $sender_address_data in subsequent conditions in the ACL statement.
644 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
646 15. Added a new option "connect=<time>" to callout options, to set a different
649 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
650 was its contents. (It was OK if the option was not defined at all.)
652 17. A "Completed" log line is now written for messages that are removed from
653 the spool by the -Mrm option.
655 18. New variables $sender_verify_failure and $recipient_verify_failure contain
656 information about exactly what failed.
658 19. Added -dd to debug only the daemon process.
660 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
661 handles timeouts, both on the server side and network timeouts. Renamed the
662 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
664 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
665 It is now set to "smtps".
667 22. $host_address is now set to the target address during the checking of
670 23. When checking ignore_target_hosts for an ipliteral router, no host name was
671 being passed; this would have caused $sender_host_name to have been used if
672 matching the list had actually called for a host name (not very likely,
673 since this list is usually IP addresses). A host name is now passed as
676 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
677 code that specifies a non-restarting handler (typically sigaction() in
678 modern systems) in an attempt to fix a rare and obscure crash bug.
680 25. Narrowed the window for a race in the daemon that could cause it to ignore
681 SIGCHLD signals. This is not a major problem, because they are used only to
682 wake it up if nothing else does.
684 26. A malformed maildirsize file could cause Exim to calculate negative values
685 for the mailbox size or file count. Odd effects could occur as a result.
686 The maildirsize information is now recalculated if the size or filecount
689 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
690 support for a long time. Removed HAVE_SYS_VFS_H.
692 28. Installed the latest version of exipick from John Jetmore.
694 29. In an address list, if the pattern was not a regular expression, an empty
695 subject address (from a bounce message) matched only if the pattern was an
696 empty string. Non-empty patterns were not even tested. This was the wrong
697 because it is perfectly reasonable to use an empty address as part of a
698 database query. An empty address is now tested by patterns that are
699 lookups. However, all the other forms of pattern expect the subject to
700 contain a local part and a domain, and therefore, for them, an empty
701 address still always fails if the pattern is not itself empty.
703 30. Exim went into a mad DNS loop when attempting to do a callout where the
704 host was specified on an smtp transport, and looking it up yielded more
707 31. Re-factored the code for checking spool and log partition space into a
708 function that finds that data and another that does the check. The former
709 is then used to implement four new variables: $spool_space, $log_space,
710 $spool_inodes, and $log_inodes.
712 32. The RFC2047 encoding function was originally intended for short strings
713 such as real names; it was not keeping to the 75-character limit for
714 encoded words that the RFC imposes. It now respects the limit, and
715 generates multiple encoded words if necessary. To be on the safe side, I
716 have increased the buffer size for the ${rfc2047: expansion operator from
719 33. It is now permitted to omit both strings after an "if" condition; if the
720 condition is true, the result is "true". As before, when the second string
721 is omitted, a false condition yields an empty string. This makes it less
722 cumbersome to write custom ACL and router conditions.
724 34. Failure to deliver a bounce message always caused it to be frozen, even if
725 there was an errors_to setting on the router. The errors_to setting is now
728 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
729 canonical form (fully expanded) before being placed in
730 $sender_host_address.
732 36. The table in the code that translates DNS record types into text (T_A to
733 "A" for instance) was missing entries for NS and CNAME. It is just possible
734 that this could have caused confusion if both these types were looked up
735 for the same domain, because the text type is used as part of Exim's
736 per-process caching. But the chance of anyone hitting this buglet seems
739 37. The dnsdb lookup has been extended in a number of ways.
741 (1) There is a new type, "zns", which walks up the domain tree until it
742 finds some nameserver records. It should be used with care.
744 (2) There is a new type, "mxh", which is like "mx" except that it returns
745 just the host names, not the priorities.
747 (3) It is now possible to give a list of domains (or IP addresses) to be
748 looked up. The behaviour when one of the lookups defers can be
749 controlled by a keyword.
751 (4) It is now possible to specify the separator character for use when
752 multiple records are returned.
754 38. The dnslists ACL condition has been extended: it is now possible to supply
755 a list of IP addresses and/or domains to be looked up in a particular DNS
758 39. Added log_selector=+queue_time_overall.
760 40. When running the queue in the test harness, wait just a tad after forking a
761 delivery process, to get repeatability of debugging output.
763 41. Include certificate and key file names in error message when GnuTLS fails
764 to set them up, because the GnuTLS error message doesn't include the name
765 of the failing file when there is a problem reading it.
767 42. Allow both -bf and -bF in the same test run.
769 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
771 44. The "Exiscan patch" is now merged into the mainline Exim source.
773 45. Sometimes the final signoff response after QUIT could fail to get
774 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
775 < 0 before doing a fflush(). This bug looks as though it goes back to the
776 introduction of TLS in release 3.20, but "sometimes" must have been rare
777 because the tests only now provoked it.
779 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
780 (this can affect the format of dates).
782 47. exim_tidydb, when checking for the continued existence of a message for
783 which it has found a message-specific retry record, was not finding
784 messages that were in split spool directories. Consequently, it was
785 deleting retry records that should have stayed in existence.
787 48. Steve fixed some bugs in eximstats.
789 49. The SPA authentication driver was not abandoning authentication and moving
790 on to the next authenticator when an expansion was forced to fail,
791 contradicting the general specification for all authenticators. Instead it
792 was generating a temporary error. It now behaves as specified.
794 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
795 (the order specifies the preference for clients). The order is now AES256,
796 AES128, 3DES, ARCFOUR128.
798 51. Small patch to Sieve code - explicitly set From: when generating an
801 52. Exim crashed if a remote delivery caused a very long error message to be
802 recorded - for instance if somebody sent an entire SpamAssassin report back
803 as a large number of 550 error lines. This bug was coincidentally fixed by
804 increasing the size of one of Exim's internal buffers (big_buffer) that
805 happened as part of the Exiscan merge. However, to be on the safe side, I
806 have made the code more robust (and fixed the comments that describe what
809 53. Now that there can be additional text after "Completed" in log lines (if
810 the queue_time_overall log selector is set), a one-byte patch to exigrep
811 was needed to allow it to recognize "Completed" as not the last thing in
814 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
815 patch that reportedly fixes this has been added. I am not expert enough to
816 create a test for it. This is what the patch creator wrote:
818 "I found a little strange behaviour of ldap code when working with
819 Windows 2003 AD Domain, where users was placed in more than one
820 Organization Units. When I tried to give exim partial DN, the exit code
821 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
822 But simultaneously result of request was absolutely normal ldap result,
823 so I produce this patch..."
825 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
826 REFERENCE, so I have modified the code to exclude the patch when that macro
829 55. Some experimental protocols are using DNS PTR records for new purposes. The
830 keys for these records are domain names, not reversed IP addresses. The
831 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
832 leaves it alone. Component reversal etc. now happens only for IP addresses.
835 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
837 57. Double the size of the debug message buffer (to 2048) so that more of very
838 long debug lines gets shown.
840 58. The exicyclog utility now does better if the number of log files to keep
841 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
843 59. Two changes related to the smtp_active_hostname option:
845 (1) $smtp_active_hostname is now available as a variable.
846 (2) The default for smtp_banner uses $smtp_active_hostname instead
847 of $primary_hostname.
849 60. The host_aton() function is supposed to be passed a string that is known
850 to be a valid IP address. However, in the case of IPv6 addresses, it was
851 not checking this. This is a hostage to fortune. Exim now panics and dies
852 if the condition is not met. A case was found where this could be provoked
853 from a dnsdb PTR lookup with an IPv6 address that had more than 8
854 components; fortuitously, this particular loophole had already been fixed
855 by change 4.50/55 above.
857 If there are any other similar loopholes, the new check in host_aton()
858 itself should stop them being exploited. The report I received stated that
859 data on the command line could provoke the exploit when Exim was running as
860 exim, but did not say which command line option was involved. All I could
861 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
865 61. There was a buffer overflow vulnerability in the SPA authentication code
866 (which came originally from the Samba project). I have added a test to the
867 spa_base64_to_bits() function which I hope fixes it.
870 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
871 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
873 63. The daemon start-up calls getloadavg() while still root for those OS that
874 need the first call to be done as root, but it missed one case: when
875 deliver_queue_load_max is set with deliver_drop_privilege. This is
876 necessary for the benefit of the queue runner, because there is no re-exec
877 when deliver_drop_privilege is set.
879 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
882 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
883 was in use, was not putting the data itself into the right store pool;
884 consequently, it could be overwritten for a subsequent message in the same
885 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
888 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
890 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
891 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
892 the function were treating the return as a boolean value, which happened to
893 work because 0=false and not-0=true, but is not correct code.
895 68. The host_aton() function was not handling scoped IPv6 addresses (those
896 with, for example, "%eth0" on the end) correctly.
898 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
899 negated items (that is, ~something) in unsigned ints. Some compilers
900 apparently mutter when there is no cast.
902 70. If an address verification called from an ACL failed, and did not produce a
903 user-specific message (i.e. there was only a "system" message), nothing was
904 put in $acl_verify_message. In this situation, it now puts the system
907 71. Change 4.23/11 added synchronization checking at the start of an SMTP
908 session; change 4.31/43 added the unwanted input to the log line - except
909 that it did not do this in the start of session case. It now does.
911 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
912 This is wrong and can cause the other end to generate a synchronization
913 error if it is another Exim or anything else that does the synchronization
914 check. A QUIT command is no longer sent after a timeout.
916 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
919 74. The defer_ok option of callout verification was not working if it was used
920 when verifying addresses in header lines, that is, for this case:
922 verify = header_sender/callout=defer_ok
924 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
925 those file descriptors could be used for SMTP connections. If anything
926 wrote to stderr (the example that came up was "warn" in embedded Perl), it
927 could be sent to the SMTP client, causing chaos. The daemon now opens
928 stdin, stdout, and stderr to /dev/null when it puts itself into the
931 76. Arrange for output from Perl's "warn" command to be written to Exim's main
932 log by default. The user can override this with suitable Perl magic.
934 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
935 the log message when discard triggers, was not working for the DATA ACL or
936 for the non-SMTP ACL.
938 78. Error message wording change in sieve.c.
940 79. If smtp_accept_max_per_host was set, the number of connections could be
941 restricted to fewer than expected, because the daemon was trying to set up
942 a new connection before checking whether the processes handling previous
943 connections had finished. The check for completed processes is now done
944 earlier. On busy systems, this bug wouldn't be noticed because something
945 else would have woken the daemon, and it would have reaped the completed
948 80. If a message was submitted locally by a user whose login name contained one
949 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
950 It caused a spool format error. I have fixed the spool reading code. A
951 related problem was that the "from" clause in the Received: line became
952 illegal because of the space(s). It is now covered by ${quote_local_part.
954 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
957 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
959 83. Patch from Sieve maintainer to fix unterminated string problem in
962 84. Some minor changes to the Linux configuration files to help with other
963 OS variants using glibc.
965 85. One more patch for Sieve to update vacation handling to latest spec.
968 ----------------------------------------------------
969 See the note above about the 4.44 and 4.50 releases.
970 ----------------------------------------------------
976 1. Change 4.43/35 introduced a bug that caused file counts to be
977 incorrectly computed when quota_filecount was set in an appendfile
980 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
981 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
984 3. Give more explanation in the error message when the command for a transport
985 filter fails to execute.
987 4. There are several places where Exim runs a non-Exim command in a
988 subprocess. The SIGUSR1 signal should be disabled for these processes. This
989 was being done only for the command run by the queryprogram router. It is
990 now done for all such subprocesses. The other cases are: ${run, transport
991 filters, and the commands run by the lmtp and pipe transports.
993 5. Some older OS have a limit of 256 on the maximum number of file
994 descriptors. Exim was using setrlimit() to set 1000 as a large value
995 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
996 systems. I've change it so that if it can't get 1000, it tries for 256.
998 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
999 was an oversight, and furthermore, ever since the addition of extra
1000 controls (e.g. 4.43/32), the checks on when to allow different forms of
1001 "control" were broken. There should now be diagnostics for all cases when a
1002 control that does not make sense is encountered.
1004 7. $recipients is now available in the predata ACL (oversight).
1006 8. Tidy the search cache before the fork to do a delivery from a message
1007 received from the command line. Otherwise the child will trigger a lookup
1008 failure and thereby defer the delivery if it tries to use (for example) a
1009 cached ldap connection that the parent has called unbind on.
1011 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1012 of $address_data from the recipient verification was clobbered by the
1013 sender verification.
1015 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1016 was its contents. (It was OK if the option was not defined at all.)
1018 11. A "Completed" log line is now written for messages that are removed from
1019 the spool by the -Mrm option.
1021 12. $host_address is now set to the target address during the checking of
1022 ignore_target_hosts.
1024 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1025 being passed; this would have caused $sender_host_name to have been used if
1026 matching the list had actually called for a host name (not very likely,
1027 since this list is usually IP addresses). A host name is now passed as
1030 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1031 code that specifies a non-restarting handler (typically sigaction() in
1032 modern systems) in an attempt to fix a rare and obscure crash bug.
1034 15. Narrowed the window for a race in the daemon that could cause it to ignore
1035 SIGCHLD signals. This is not a major problem, because they are used only to
1036 wake it up if nothing else does.
1038 16. A malformed maildirsize file could cause Exim to calculate negative values
1039 for the mailbox size or file count. Odd effects could occur as a result.
1040 The maildirsize information is now recalculated if the size or filecount
1043 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1044 support for a long time. Removed HAVE_SYS_VFS_H.
1046 18. Updated exipick to current release from John Jetmore.
1048 19. Allow an empty sender to be matched against a lookup in an address list.
1049 Previously the only cases considered were a regular expression, or an
1052 20. Exim went into a mad DNS lookup loop when doing a callout where the
1053 host was specified on the transport, if the DNS lookup yielded more than
1056 21. The RFC2047 encoding function was originally intended for short strings
1057 such as real names; it was not keeping to the 75-character limit for
1058 encoded words that the RFC imposes. It now respects the limit, and
1059 generates multiple encoded words if necessary. To be on the safe side, I
1060 have increased the buffer size for the ${rfc2047: expansion operator from
1063 22. Failure to deliver a bounce message always caused it to be frozen, even if
1064 there was an errors_to setting on the router. The errors_to setting is now
1067 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1068 canonical form (fully expanded) before being placed in
1069 $sender_host_address.
1071 24. Updated eximstats to version 1.33
1073 25. Include certificate and key file names in error message when GnuTLS fails
1074 to set them up, because the GnuTLS error message doesn't include the name
1075 of the failing file when there is a problem reading it.
1077 26. Expand error message when OpenSSL has problems setting up cert/key files.
1080 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1081 (this can affect the format of dates).
1083 28. exim_tidydb, when checking for the continued existence of a message for
1084 which it has found a message-specific retry record, was not finding
1085 messages that were in split spool directories. Consequently, it was
1086 deleting retry records that should have stayed in existence.
1088 29. eximstats updated to version 1.35
1089 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1090 1.35 - bugfix such that pie charts by volume are generated correctly
1092 30. The SPA authentication driver was not abandoning authentication and moving
1093 on to the next authenticator when an expansion was forced to fail,
1094 contradicting the general specification for all authenticators. Instead it
1095 was generating a temporary error. It now behaves as specified.
1097 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1098 (the order specifies the preference for clients). The order is now AES256,
1099 AES128, 3DES, ARCFOUR128.
1101 31. Small patch to Sieve code - explicitly set From: when generating an
1104 32. Exim crashed if a remote delivery caused a very long error message to be
1105 recorded - for instance if somebody sent an entire SpamAssassin report back
1106 as a large number of 550 error lines. This bug was coincidentally fixed by
1107 increasing the size of one of Exim's internal buffers (big_buffer) that
1108 happened as part of the Exiscan merge. However, to be on the safe side, I
1109 have made the code more robust (and fixed the comments that describe what
1112 33. Some experimental protocols are using DNS PTR records for new purposes. The
1113 keys for these records are domain names, not reversed IP addresses. The
1114 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1115 leaves it alone. Component reversal etc. now happens only for IP addresses.
1118 34. The host_aton() function is supposed to be passed a string that is known
1119 to be a valid IP address. However, in the case of IPv6 addresses, it was
1120 not checking this. This is a hostage to fortune. Exim now panics and dies
1121 if the condition is not met. A case was found where this could be provoked
1122 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1123 components; fortuitously, this particular loophole had already been fixed
1124 by change 4.50/55 or 4.44/33 above.
1126 If there are any other similar loopholes, the new check in host_aton()
1127 itself should stop them being exploited. The report I received stated that
1128 data on the command line could provoke the exploit when Exim was running as
1129 exim, but did not say which command line option was involved. All I could
1130 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1131 running as the user.
1134 35. There was a buffer overflow vulnerability in the SPA authentication code
1135 (which came originally from the Samba project). I have added a test to the
1136 spa_base64_to_bits() function which I hope fixes it.
1139 36. The daemon start-up calls getloadavg() while still root for those OS that
1140 need the first call to be done as root, but it missed one case: when
1141 deliver_queue_load_max is set with deliver_drop_privilege. This is
1142 necessary for the benefit of the queue runner, because there is no re-exec
1143 when deliver_drop_privilege is set.
1145 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1146 was in use, was not putting the data itself into the right store pool;
1147 consequently, it could be overwritten for a subsequent message in the same
1148 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1151 38. Sometimes the final signoff response after QUIT could fail to get
1152 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1153 < 0 before doing a fflush(). This bug looks as though it goes back to the
1154 introduction of TLS in release 3.20, but "sometimes" must have been rare
1155 because the tests only now provoked it.
1161 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1162 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1163 Now it returns an integer. A number of calls were still expecting a T/F
1164 return. Fortuitously, in all cases, the tests worked in OK situations,
1165 which is the norm. However, things would have gone wrong on any write
1166 failures on the smtp file descriptor. This function is used when sending
1167 messages over SMTP and also when doing verify callouts.
1169 2. When Exim is called to do synchronous delivery of a locally submitted
1170 message (the -odf or -odi options), it no longer closes stderr before doing
1173 3. Implemented the mua_wrapper option.
1175 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1177 5. Implemented the functions header_remove(), header_testname(),
1178 header_add_at_position(), and receive_remove_recipient(), and exported them
1181 6. If an ACL "warn" statement specified the addition of headers, Exim already
1182 inserted X-ACL-Warn: at the start if there was no header name. However, it
1183 was not making this test for the second and subsequent header lines if
1184 there were newlines in the string. This meant that an invalid header could
1185 be inserted if Exim was badly configured.
1187 7. Allow an ACL "warn" statement to add header lines at the start or after all
1188 the Received: headers, as well as at the end.
1190 8. Added the rcpt_4xx retry error code.
1192 9. Added postmaster_mailfrom=xxx to callout verification option.
1194 10. Added mailfrom=xxxx to the callout verification option, for verify=
1197 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1198 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1200 12. Inserted some casts to stop certain compilers warning when using pointer
1201 differences as field lengths or precisions in printf-type calls (mostly
1202 affecting debugging statements).
1204 13. Added optional readline() support for -be (dynamically loaded).
1206 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1207 same clock tick as a message's arrival, so that its received time was the
1208 same as the "first fail" time on the retry record, and that message
1209 remained on the queue past the ultimate address timeout, every queue runner
1210 would try a delivery (because it was past the ultimate address timeout) but
1211 after another failure, the ultimate address timeout, which should have then
1212 bounced the address, did not kick in. This was a "< instead of <=" error;
1213 in most cases the first failure would have been in the next clock tick
1214 after the received time, and all would be well.
1216 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1217 being recognized when the domain list was tested by the match_domain
1218 condition in an expansion string.
1220 16. Added the ${str2b64: operator.
1222 17. Exim was always calling setrlimit() to set a large limit for the number of
1223 processes, without checking whether the existing limit was already
1224 adequate. (It did check for the limit on file descriptors.) Furthermore,
1225 errors from getrlimit() and setrlimit() were being ignored. Now they are
1226 logged to the main and panic logs, but Exim does carry on, to try to do its
1227 job under whatever limits there are.
1229 18. Imported PCRE 5.0.
1231 19. Trivial typo in log message " temporarily refused connection" (the leading
1234 20. If the log selector return_path_on_delivery was set and an address was
1235 redirected to /dev/null, the delivery process crashed because it assumed
1236 that a return path would always be set for a "successful" delivery. In this
1237 case, the whole delivery is bypassed as an optimization, and therefore no
1240 21. Internal re-arrangement: the function for sending a challenge and reading
1241 a response while authentication was assuming a zero-terminated challenge
1242 string. It's now changed to take a pointer and a length, to allow for
1243 binary data in such strings.
1245 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1247 23. Exim was not respecting finduser_retries when seeking the login of the
1248 uid under which it was called; it was always trying 10 times. (The default
1249 setting of finduser_retries is zero.) Also, it was sleeping after the final
1250 failure, which is pointless.
1252 24. Implemented tls_on_connect_ports.
1254 25. Implemented acl_smtp_predata.
1256 26. If the domain in control=submission is set empty, Exim assumes that the
1257 authenticated id is a complete email address when it generates From: or
1258 Sender: header lines.
1260 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1261 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1262 chown and chgrp in /bin and hostname in /usr/bin.
1264 28. Exim was keeping the "process log" file open after each use, just as it
1265 does for the main log. This opens the possibility of it remaining open for
1266 long periods when the USR1 signal hits a daemon. Occasional processlog
1267 errors were reported, that could have been caused by this. Anyway, it seems
1268 much more sensible not to leave this file open at all, so that is what now
1271 29. The long-running daemon process does not normally write to the log once it
1272 has entered its main loop, and it closes the log before doing so. This is
1273 so that log files can straightforwardly be renamed and moved. However,
1274 there are a couple of unusual error situations where the daemon does write
1275 log entries, and I had neglected to close the log afterwards.
1277 30. The text of an SMTP error response that was received during a remote
1278 delivery was being truncated at 512 bytes. This is too short for some of
1279 the long messages that one sometimes sees. I've increased the limit to
1282 31. It is now possible to make retry rules that apply only when a message has a
1283 specific sender, in particular, an empty sender.
1285 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1286 it possible to be selective about when SMTP synchronization is enforced.
1288 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1290 32. Implemented hosts_connection_nolog.
1292 33. Added an ACL for QUIT.
1294 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1297 35. Added mailbox_size and mailbox_filecount to appendfile.
1299 36. Added control = no_multiline_responses to ACLs.
1301 37. There was a bug in the logic of the code that waits for the clock to tick
1302 in the case where the clock went backwards by a substantial amount such
1303 that the microsecond fraction of "now" was more than the microsecond
1304 fraction of "then" (but the whole seconds number was less).
1306 38. Added support for the libradius Radius client library this is found on
1307 FreeBSD (previously only the radiusclient library was supported).
1313 1. When certain lookups returned multiple values in the form name=value, the
1314 quoting of the values was not always being done properly. Specifically:
1315 (a) If the value started with a double quote, but contained no whitespace,
1317 (b) If the value contained whitespace other than a space character (i.e.
1318 tabs or newlines or carriage returns) it was not quoted.
1319 This fix has been applied to the mysql and pgsql lookups by writing a
1320 separate quoting function and calling it from the lookup code. The fix
1321 should probably also be applied to nisplus, ibase and oracle lookups, but
1322 since I cannot test any of those, I have not disturbed their existing code.
1324 2. A hit in the callout cache for a specific address caused a log line with no
1325 reason for rejecting RCPT. Now it says "Previous (cached) callout
1326 verification failure".
1328 3. There was an off-by-one bug in the queryprogram router. An over-long
1329 return line was truncated at 256 instead of 255 characters, thereby
1330 overflowing its buffer with the terminating zero. As well as fixing this, I
1331 have increased the buffer size to 1024 (and made a note to document this).
1333 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1334 when Exim is waiting for an SMTP response from a remote server, Exim
1335 restarts its select() call on the socket, thereby resetting its timeout.
1336 This is not a problem when such interrupts are rare. Somebody set up a cron
1337 job to run exiwhat every 2 minutes, which is less than the normal select()
1338 timeout (5 or 10 minutes). This meant that the select() timeout never
1339 kicked in because it was always reset. I have fixed this by comparing the
1340 time when an interrupt arrives with the time at the start of the first call
1341 to select(). If more time than the timeout has elapsed, the interrupt is
1342 treated as a timeout.
1344 5. Some internal re-factoring in preparation for the addition of Sieve
1345 extensions (by MH). In particular, the "personal" test is moved to a
1346 separate function, and given an option for scanning Cc: and Bcc: (which is
1347 not set for Exim filters).
1349 6. When Exim created an email address using the login of the caller as the
1350 local part (e.g. when creating a From: or Sender: header line), it was not
1351 quoting the local part when it contained special characters such as @.
1353 7. Installed new OpenBSD configuration files.
1355 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1356 try to make them clearer.
1358 9. Callout options, other than the timeout value, were being ignored when
1359 verifying sender addresses in header lines. For example, when using
1361 verify = header_sender/callout=no_cache
1363 the cache was (incorrectly) being used.
1365 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1366 only the Cygwin environment.
1368 11. When return_path_on_delivery was set as a log selector, if different remote
1369 addresses in the same message used different return paths and parallel
1370 remote delivery occurred, the wrong values would sometimes be logged.
1371 (Whenever a remote delivery process finished, the return path value from
1372 the most recently started remote delivery process was logged.)
1374 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1375 lines when AUTH and/or TLS are in use. This is the "received protocol"
1376 field. Exim used to use "asmtp" for authenticated SMTP, without any
1377 indication (in the protocol name) for TLS use. Now it follows the RFC and
1378 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1379 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1380 names appear in log lines as well as in Received: header lines.
1382 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1383 extensions, and comparison tests, and to fix some bugs.
1385 14. Changes to the "personal" filter test:
1387 (1) The test was buggy in that it was just doing the equivalent of
1388 "contains" tests on header lines. For example, if a user's address was
1389 anne@some.where, the "personal" test would incorrectly be true for
1391 To: susanne@some.where
1393 This test is now done by extracting each address from the header in turn,
1394 and checking the entire address. Other tests that are part of "personal"
1395 are now done using regular expressions (for example, to check local parts
1396 of addresses in From: header lines).
1398 (2) The list of non-personal local parts in From: addresses has been
1399 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1400 taken from the Sieve specification recommendations.
1402 (3) If the message contains any header line starting with "List-" it is
1403 treated as non-personal.
1405 (4) The test for "circular" in the Subject: header line has been removed
1406 because it now seems ill-conceived.
1408 15. Minor typos in src/EDITME comments corrected.
1410 16. Installed latest exipick from John Jetmore.
1412 17. If headers_add on a router specified a text string that was too long for
1413 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1414 of string_sprintf() is now avoided.
1416 18. $message_body_size was not set (it was always zero) when running the DATA
1417 ACL and the local_scan() function.
1419 19. For the "mail" command in an Exim filter, no default was being set for
1420 the once_repeat time, causing a random time value to be used if "once" was
1421 specified. (If the value happened to be <= 0, no repeat happened.) The
1422 default is now 0s, meaning "never repeat". The "vacation" command was OK
1423 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1424 (I found it when inspecting the code).
1426 20. There is now an overall timeout for performing a callout verification. It
1427 defaults to 4 times the callout timeout, which applies to individual SMTP
1428 commands during the callout. The overall timeout applies when there is more
1429 than one host that can be tried. The timeout is checked before trying the
1430 next host. This prevents very long delays if there are a large number of
1431 hosts and all are timing out (e.g. when the network connections are timing
1432 out). The value of the overall timeout can be changed by specifying an
1433 additional sub-option for "callout", called "maxwait". For example:
1435 verify = sender/callout=5s,maxwait=20s
1437 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1438 to the end before writing, but this should make it even safer).
1440 22. Exim was forgetting that it had advertised PIPELINING for the second and
1441 subsequent messages on an SMTP connection. It was also not resetting its
1442 memory on STARTTLS and an internal HELO.
1444 23. When Exim logs an SMTP synchronization error within a session, it now
1445 records whether PIPELINING has been advertised or not.
1447 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1448 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1449 rather than long int.
1451 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1453 26. Added the never_mail option to autoreply.
1459 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1460 crash if the getsockname() call failed; this can happen if a connection is
1461 closed very soon after it is established. The problem was simply in the
1462 order in which certain operations were done, causing Exim to try to write
1463 to the SMTP stream before it had set up the file descriptor. The bug has
1464 been fixed by making things happen in the correct order.
1470 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1471 before the connection was closed, thus losing the rejection response.
1473 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1474 some early Solaris releases, but causes trouble in current releases where
1475 socklen_t is defined.
1477 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1480 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1483 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1484 that releasing the top of it at the end releases what was used for sub-
1485 expansions (unless the block got too big). However, discard this block if
1486 the first thing is a variable or header, so that we can use its block when
1487 it is dynamic (useful for very large $message_headers, for example).
1489 6. Lookups now cache *every* query, not just the most recent. A new, separate
1490 store pool is used for this. It can be recovered when all lookup caches are
1491 flushed. Lookups now release memory at the end of their result strings.
1492 This has involved some general refactoring of the lookup sources.
1494 7. Some code has been added to the store_xxx() functions to reduce the amount
1495 of flapping under certain conditions.
1497 8. log_incoming_interface used to affect only the <= reception log lines. Now
1498 it causes the local interface and port to be added to several more SMTP log
1499 lines, for example "SMTP connection from", and rejection lines.
1501 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1503 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1505 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1506 could be overwritten at the end of the current message (or the start of a
1507 new message if it was set in a HELO ACL). The value is now preserved for
1508 the duration of the SMTP connection.
1510 12. If a transport had a headers_rewrite setting, and a matching header line
1511 contained an unqualified address, that address was qualified, even if it
1512 did not match any rewriting rules. The underlying bug was that the values
1513 of the flags that permit the existence of unqualified sender and recipient
1514 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1515 non-local messages, and by -bnq for local messages) were not being
1516 preserved with the message after it was received.
1518 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1519 "next input=" as part of the text comprising the host identity instead of
1520 the correct text. The code was using the same buffer for two different
1521 strings. However, depending on which order the printing function evaluated
1522 its arguments, the bug did not always show up. Under Linux, for example, my
1523 test suite worked just fine.
1525 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1526 doesn't work with some older versions of Perl. It has been changed to "my",
1527 which in any case is probably the better facility to use.
1529 15. A really picky compiler found some instances of statements for creating
1530 error messages that either had too many or two few arguments for the format
1533 16. The size of the buffer for calls to the DNS resolver has been increased
1534 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1535 for addresses that have a lot of PTR records. This alleviates a problem; it
1536 does not fully solve it.
1538 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1539 buffer now truncates the list and logs the incident, which is the same
1540 action as happens when Exim is looking up a host name and its aliases.
1541 Previously in this situation something unpredictable would happen;
1542 sometimes it was "internal error: store_reset failed".
1544 18. If a server dropped the connection unexpectedly when an Exim client was
1545 using GnuTLS and trying to read a response, the client delivery process
1546 crashed while trying to generate an error log message.
1548 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1549 string, for example:
1551 warn message = H1: something\nH2: something
1553 the text was added as a single header line from Exim's point of view
1554 though it ended up OK in the delivered message. However, searching for the
1555 second and subsequent header lines using $h_h2: did not work. This has been
1556 fixed. Similarly, if a system filter added multiple headers in this way,
1557 the routers could not see them.
1559 20. Expanded the error message when iplsearch is called with an invalid key to
1560 suggest using net-iplsearch in a host list.
1562 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1563 ACLs are no longer actually imposed (and a message to that effect is
1566 22. If a "gecos" field in a passwd entry contained escaped characters, in
1567 particular, if it contained a \" sequence, Exim got it wrong when building
1568 a From: or a Sender: header from that name. A second bug also caused
1569 incorrect handling when an unquoted " was present following a character
1570 that needed quoting.
1572 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1573 was not being matched caselessly.
1575 24. Arranged for all hyphens in the exim.8 source to be escaped with
1578 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1579 properly. Recipient callout cache records were still being keyed to include
1580 the sender, even when use_sender was set false. This led to far more
1581 callouts that were necessary. The sender is no longer included in the key
1582 when use_sender is false.
1584 26. Added "control = submission" modifier to ACLs.
1586 27. Added the ${base62d: operator to decode base 62 numbers.
1588 28. dnsdb lookups can now access SRV records.
1590 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1591 the configuration file.
1593 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1594 (-v) mode. This makes the output for a verbose queue run more intelligible.
1596 31. Added a use_postmaster feature to recipient callouts.
1598 32. Added the $body_zerocount variable, containing the number of binary zero
1599 bytes in the message body.
1601 33. The time of last modification of the "new" subdirectory is now used as the
1602 "mailbox time last read" when there is a quota error for a maildir
1605 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1607 35. Added +ignore_unknown as a special item in host lists.
1609 36. Code for decoding IPv6 addresses in host lists is now included, even if
1610 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1611 address was recognized as an IP address, but was then not correctly decoded
1612 into binary, causing unexpected and incorrect effects when compared with
1619 1. Very minor rewording of debugging text in manualroute to say "list of
1620 hosts" instead of "hostlist".
1622 2. If verify=header_syntax was set, and a header line with an unqualified
1623 address (no domain) and a large number of spaces between the end of the
1624 name and the colon was received, the reception process suffered a buffer
1625 overflow, and (when I tested it) crashed. This was caused by some obsolete
1626 code that should have been removed. The fix is to remove it!
1628 3. When running in the test harness, delay a bit after writing a bounce
1629 message to get a bit more predictability in the log output.
1631 4. Added a call to search_tidyup() just before forking a reception process. In
1632 theory, someone could use a lookup in the expansion of smtp_accept_max_
1633 per_host which, without the tidyup, could leave open a database connection.
1635 5. Added the variables $recipient_data and $sender_data which get set from a
1636 lookup success in an ACL "recipients" or "senders" condition, or a router
1637 "senders" option, similar to $domain_data and $local_part_data.
1639 6. Moved the writing of debug_print from before to after the "senders" test
1642 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1643 problems for message scanning, either using a data ACL, or using
1644 local_scan() because the Received: header was not generated till after they
1645 were called (in order to set the time as the time of reception completion).
1646 I have revised the way this works. The header is now generated after the
1647 body is received, but before the ACL or local_scan() are called. After they
1648 are run, the timestamp in the header is updated.
1654 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1655 before starting a queue runner without re-exec. This happened only when
1656 deliver_drop_privilege was set or when the Exim user was set to root. The
1657 effect of the bug was that timeouts during subsequent deliveries caused
1658 crashes instead of being properly handled. The handler is now left at its
1659 default (and expected) setting.
1661 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1662 message, again when deliver_drop_privilege is set or Exim is run as root.
1663 The bug described in (1) was not present in this case, but the tidying up
1664 of the other signals was missing. I have made the two cases consistent.
1666 3. The ignore_target_hosts setting on a manualroute router was being ignored
1667 for hosts that were looked up using the /MX notation.
1669 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1672 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1673 operated on the sender address. After changing the $sender_address to <>
1674 for the sender address verify, Exim was re-instated it as the original
1675 (before rewriting) address, but remembering that it had rewritten it, so it
1676 wasn't rewriting it again. This bug also had the effect of breaking the
1677 sender address verification caching when the sender address was rewritten.
1679 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1680 This has been changed so that if the ip literal address matches
1681 ignore_target_hosts, the router declines.
1683 7. Added expansion conditions match_domain, match_address, and match_local_
1684 part (NOT match_host).
1686 8. The placeholder for the Received: header didn't have a length field set.
1688 9. Added code to Exim itself and to exim_lock to test for a specific race
1689 condition that could lead to file corruption when using MBX delivery. The
1690 issue is with the lockfile that is created in /tmp. If this file is removed
1691 after a process has opened it but before that process has acquired a lock,
1692 there is the potential for a second process to recreate the file and also
1693 acquire a lock. This could lead to two Exim processes writing to the file
1694 at the same time. The added code performs the same test as UW imapd; it
1695 checks after acquiring the lock that its file descriptor still refers to
1696 the same named file.
1698 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1699 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1702 11. Added the smtp_active_hostname option. If used, this will typically be made
1703 to depend on the incoming interface address. Because $interface_address is
1704 not set up until the daemon has forked a reception process, error responses
1705 that can happen earlier (such as "too many connections") no longer contain
1708 12. If an expansion in a condition on a "warn" statement fails because a lookup
1709 defers, the "warn" statement is abandoned, and the next ACL statement is
1710 processed. Previously this caused the whole ACL to be aborted.
1712 13. Added the iplsearch lookup type.
1714 14. Added ident_timeout as a log selector.
1716 15. Added tls_certificate_verified as a log selector.
1718 16. Added a global option tls_require_ciphers (compare the smtp transport
1719 option of the same name). This controls incoming TLS connections.
1721 17. I finally figured out how to make tls_require_ciphers do a similar thing
1722 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1723 before starting the TLS session.
1725 18. Tabs are now shown as \t in -bP output.
1727 19. If the log selector return_path_on_delivery was set, Exim crashed when
1728 bouncing a message because it had too many Received: header lines.
1730 20. If two routers both had headers_remove settings, and the first one included
1731 a superfluous trailing colon, the final name in the first list and the
1732 first name in the second list were incorrectly joined into one item (with a
1733 colon in the middle).
1739 1. Added -C and -D options to the exinext utility, mainly to make it easier
1740 to include in the automated testing, but these could be helpful when
1741 multiple configurations are in use.
1743 2. The exinext utility was not formatting the output nicely when there was
1744 an alternate port involved in the retry record key, nor when there was a
1745 message id as well (for retries that were specific to a specific message
1746 and a specific host). It was also confused by IPv6 addresses, because of
1747 the additional colons they contain. I have fixed the IPv4 problem, and
1748 patched it up to do a reasonable job for IPv6.
1750 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1751 delivery, the log line now contains "pipelined" if PIPELINING was used.
1753 4. An SMTP transport process used to panic and die if the bind() call to set
1754 an explicit outgoing interface failed. This has been changed; it is now
1755 treated in the same way as a connect() failure.
1757 5. A reference to $sender_host_name in the part of a conditional expansion
1758 that was being skipped was still causing a DNS lookup. This no longer
1761 6. The def: expansion condition was not recognizing references to header lines
1762 that used bh_ and bheader_.
1764 7. Added the _cache feature to named lists.
1766 8. The code for checking quota_filecount in the appendfile transport was
1767 allowing one more file than it should have been.
1769 9. For compatibility with Sendmail, the command line option
1777 and sets the incoming protocol and host name (for trusted callers). The
1778 host name and its colon can be omitted when only the protocol is to be set.
1779 Note the Exim already has two private options, -pd and -ps, that refer to
1780 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1781 "s", but I don't think that's a major issue.
1783 10. A number of refactoring changes to the code, none of which should affect
1786 (a) The number of logging options was getting close to filling up the
1787 32-bit word that was used as a bit map. I have split them into two classes:
1788 those that are passed in the argument to log_write(), and those that are
1789 only ever tested independently outside of that function. These are now in
1790 separate 32-bit words, so there is plenty of room for expansion again.
1791 There is no change in the user interface or the logging behaviour.
1793 (b) When building, for example, log lines, the code previously used a
1794 macro that called string_cat() twice, in order to add two strings. This is
1795 not really sufficiently general. Furthermore, there was one instance where
1796 it was actually wrong because one of the argument was used twice, and in
1797 one call a function was used. (As it happened, calling the function twice
1798 did not affect the overall behaviour.) The macro has been replaced by a
1799 function that can join an arbitrary number of extra strings onto a growing
1802 (c) The code for expansion conditions now uses a table and a binary chop
1803 instead of a serial search (which was left over from when there were very
1804 few conditions). Also, it now recognizes conditions like "pam" even when
1805 the relevant support is not compiled in: a suitably worded error message is
1806 given if an attempt is made to use such a condition.
1808 11. Added ${time_interval:xxxxx}.
1810 12. A bug was causing one of the ddress fields not to be passed back correctly
1811 from remote delivery subprocesses. The field in question was not being
1812 subsequently used, so this caused to problems in practice.
1814 13. Added new log selectors queue_time and deliver_time.
1816 14. Might have fixed a bug in maildirsizefile handling that threw up
1817 "unexpected character" debug warnings, and recalculated the data
1818 unnecessarily. In any case, I expanded the warning message to give more
1821 15. Added the message "Restricted characters in address" to the statements in
1822 the default ACL that block characters like @ and % in local parts.
1824 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1825 Three changes have been made:
1827 (a) There was a serious bug; a negative response to MAIL caused the whole
1828 recipient domain to be cached as invalid, thereby blocking all messages
1829 to all local parts at the same domain, from all senders. This bug has
1830 been fixed. The domain is no longer cached after a negative response to
1831 MAIL if the sender used is not empty.
1833 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1836 (c) A new callout option, "use_sender" has been added for people who want
1837 the modified behaviour.
1843 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1846 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1847 indeed breaks things for older releases.
1849 3. Added additional logging to the case where there is a problem reading data
1850 from a filter that is running in a subprocess using a pipe, in order to
1851 try to track down a specific problem.
1853 4. Testing facility fudge: when running in the test harness and attempting
1854 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1855 getting "No route to host". Convert this to a timeout.
1857 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1860 6. Some OS don't have socklen_t but use size_t instead. This affects the
1861 fifth argument of getsockopt() amongst other things. This is now
1862 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1863 can be set for individual OS. I have set it for SunOS5, OSF1, and
1864 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1865 some earlier ones do not.
1867 7. Change 4.30/15 was not doing the test caselessly.
1869 8. The standard form for an IPv6 address literal was being rejected by address
1870 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1871 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1872 this, as well as the form without the "IPv6" on the front (but only when
1873 address literals are enabled, of course).
1875 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1877 10. Exim crashed if a message with an empty sender address specified by -f
1878 encountered a router with an errors_to setting. This could be provoked only
1879 by a command such as
1883 where an empty string was supplied; "<>" did not hit this bug.
1885 11. Installed PCRE release 4.5.
1887 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1888 remained set. It is now erased.
1890 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1891 times from message ids (which are base 36 rather than the normal 62).
1893 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1894 were being counted as actual protocol errors, and logged if the log
1895 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1896 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1897 and DATA following a set of rejected RCPTs do not count as protocol errors.
1898 In other words, Exim assumes they were pipelined, though this may not
1899 actually be the case. Of course, in all cases the client gets an
1900 appropriate error code.
1902 15. If a lookup fails in an ACL condition, a message about the failure may
1903 be available; it is used if testing the ACL cannot continue, because most
1904 such messages specify what the cause of the deferral is. However, some
1905 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1906 that caused an old message to be retained and used if a later statement
1907 caused a defer, replacing the real cause of the deferral.
1909 16. If an IP address had so many PTR records that the DNS lookup buffer
1910 was not large enough to hold them, Exim could crash while trying to process
1911 the truncated data. It now detects and logs this case.
1913 17. Further to 4.21/58, another change has been made: if (and only if) the
1914 first line of a message (the first header line) ends with CRLF, a bare LF
1915 in a subsequent header line has a space inserted after it, so as not to
1916 terminate the header.
1918 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1919 unnecessarily, used atoi() instead of strtol(), and didn't check the
1920 termination when getting file sizes from file names by regex.
1922 19. Completely re-implemented the support for maildirsize files, in the light
1923 of a number of problems with the previous contributed implementation
1924 (4.30/29). In particular:
1926 . If the quota is zero, the maildirsize file is maintained, but no quota is
1929 . If the maildir directory does not exist, it is created before any attempt
1930 to write a maildirsize file.
1932 . The quota value in the file is just a cache; if the quota is changed in
1933 the transport, the new value overrides.
1935 . A regular expression is available for excluding directories from the
1938 20. The autoreply transport checks the characters in options that define the
1939 message's headers; it allows continued headers, but it was checking with
1940 isspace() after an embedded newline instead of explicitly looking for a
1943 21. If all the "regular" hosts to which an address was routed had passed their
1944 expiry times, and had not reached their retry times, the address was
1945 bounced, even if fallback hosts were defined. Now Exim should go on to try
1948 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1949 equivalent code in the SMTP transport. Some hosts send humungous responses
1950 to HELO/EHLO, more than 1024 it seems.
1952 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1953 gives compiler warnings in some environments. I've now done it "properly",
1956 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1957 (because of problems with the built-in one) was declared to return uschar *
1958 instead of char *, causing compiler failure.
1960 25. Fixed a file descriptor leak when processing alias/forward files.
1962 26. Fixed a minor format string issue in dbfn.c.
1964 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1966 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1967 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1970 29. When a sender address is verified, it is cached, to save repeating the test
1971 when there is more than one recipient in a message. However, when the
1972 verification involves a callout, it is possible for different callout
1973 options to be set for different recipients. It is too complicated to keep
1974 track of this in the cache, so now Exim always runs a verification when a
1975 callout is required, relying on the callout cache for the optimization.
1976 The overhead is duplication of the address routing, but this should not be
1979 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
1980 to be verified with callout=postmaster, and the main callout worked but the
1981 postmaster check failed, the verification correctly failed. However, if a
1982 subsequent RCPT command asked for sender verification *without* the
1983 postmaster check, incorrect caching caused this verification also to fail,
1986 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1987 it was not caching the DNS options (qualify_single, search_parents) that
1988 were used when the lookup failed. A subsequent lookup with different
1989 options therefore always gave the same answer, though there were cases
1990 where it should not have. (Example: a "domains = !$mx_any" option on a
1991 dnslookup router: the "domains" option is always processed without any
1992 widening, but the router might have qualify_single set.) Now Exim uses the
1993 cached value only when the same options are set.
1995 32. Added John Jetmore's "exipick" utility to the distribution.
1997 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
1998 than a timeout (e.g. a certificate is required, and is not provided), an
1999 Exim server now closes the connection immediately. Previously it waited for
2000 the client to close - but if the client is SSL, it seems that they each
2001 wait for each other, leading to a delay before one of them times out.
2003 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2004 maintained 0.8.x compatibility because I don't think many are using it, and
2005 it is clearly obsolete.
2007 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2010 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2011 client certificate was expired. A simple patch fixes this, though I don't
2012 understand the full logic of why the verify callback is called multiple
2015 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2016 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2017 which causes problems with some clients (such as the Certicom SSL Plus
2018 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2019 disables the coutermeasure allowing Eudora to connect."
2021 38. Exim was not checking that a write() to a log file succeeded. This could
2022 lead to Bad Things if a log got too big, in particular if it hit a file
2023 size limit. Exim now panics and dies if it cannot write to a log file, just
2024 as it does if it cannot open a log file.
2026 39. Modified OS/Makefile-Linux so that it now contains
2028 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2030 The two -D definitions ensure that Exim is compiled with large file
2031 support, which makes it possible to handle log files that are bigger than
2034 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2035 instance) a domain was checked against a named list that involved a lookup,
2036 causing $domain_data to be set, then another domain was checked against the
2037 same list, then the first domain was re-checked, the value of $domain_data
2038 after the final check could be wrong. In particular, if the second check
2039 failed, it could be set empty. This bug probably also applied to
2042 41. The strip_trailing_dot option was not being applied to the address given
2043 with the -f command-line option.
2045 42. The code for reading a message's header from the spool was incrementing
2046 $received_count, but never initializing it. This meant that the value was
2047 incorrect (doubled) while delivering a message in the same process in which
2048 it was received. In the most common configuration of Exim, this never
2049 happens - a fresh exec is done - but it can happen when
2050 deliver_drop_privilege is set.
2052 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2053 it now includes up to 150 characters of the unexpected data in the log
2056 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2057 and building data strings. The size of both of these buffers was 10 000
2058 bytes - far larger than anybody would *ever* want, thought I. Needless to
2059 say, somebody hit the limit. I have increased the maximum line length to
2060 20 000 and the maximum data length of concatenated lines to 100 000. I have
2061 also fixed two bugs, because there was no checking on these buffers. Tsk,
2062 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2063 buffer is too small.
2065 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2066 lsearch lookups. Now it does.
2068 46. When parsing a route_list item in a manualroute router, a fixed-length
2069 buffer was used for the list of hosts. I made this 1024 bytes long,
2070 thinking that nobody would ever have a list of hosts that long. Wrong.
2071 Somebody had a whole pile of complicated expansion conditions, and the
2072 string was silently truncated, leading to an expansion error. It turns out
2073 that it is easier to change to an unlimited length (owing to other changes
2074 that have happened since this code was originally written) than to build
2075 structure for giving a limitation error. The length of the item that
2076 expands into the list of hosts is now unlimited.
2078 47. The lsearch lookup could not handle data where the length of text line was
2079 more than 4095 characters. Such lines were truncated, leading to shortened
2080 data being returned. It should now handle lines of any length.
2082 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2083 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2086 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2087 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2088 no longer happens in comment lines. A list of replacements is now placed
2089 at the head of all of the source files, except those whose only change is
2090 to replace PERL_COMMAND in the very first #! line.
2092 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2093 messages on the queue, with a bottom-up merge sort, using code contributed
2094 by Michael Haardt. This should make operations like -bp somewhat faster on
2095 large queues. It won't affect queue runners, except when queue_run_in_order
2098 51. Installed eximstats 1.31 in the distribution.
2100 52. Added support for SRV lookups to the dnslookup router.
2102 53. If an ACL referred to $message_body or $message_body_end, the value was not
2103 reset for any messages that followed in the same SMTP session.
2105 54. The store-handling optimization for building very long strings was not
2106 differentiating between the different store pools. I don't think this
2107 actually made any difference in practice, but I've tidied it.
2109 55. While running the routers to verify a sender address, $sender_address
2110 was still set to the sender address. This is wrong, because when routing to
2111 send a bounce to the sender, it would be empty. Therefore, I have changed
2112 it so that, while verifying a sender address, $sender_address is set to <>.
2113 (There is no change to what happens when verifying a recipient address.)
2115 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2116 target A or AAAA records (if not already returned) without resetting the
2117 qualify_single or search_parents options of the DNS resolver. These are
2118 inappropriate in this case because the targets of MX and SRV records must
2119 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2120 target that, when qualified, matched something in the local domain. These
2121 two options are now turned off when doing these lookups.
2123 57. It seems that at least some releases of Reiserfs (which does not have the
2124 concept of a fixed number of inodes) returns zero and not -1 for the
2125 number of available inodes. This interacted badly with check_spool_inodes,
2126 which assumed that -1 was the "no such thing" setting. What I have done is
2127 to check that the total number of inodes is greater than zero before doing
2128 the test of how many are available.
2130 58. When a "warn" ACL statement has a log_message modifier, the message is
2131 remembered, and not repeated. This is to avoid a lot of repetition when a
2132 message has many recipients that cause the same warning to be written.
2133 Howewer, Exim was preserving the list of already written lines for an
2134 entire SMTP session, which doesn't seem right. The memory is now reset if a
2135 new message is started.
2137 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2138 debugging output unless log_rewrite was also set.
2140 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2141 of (int)(handle) when we know that handle contains (void *)(-1).
2143 61. The Exim daemon panic-logs an error return when it closes the incoming
2144 connection. However "connection reset by peer" seems to be common, and
2145 isn't really an error worthy of noting specially, so that particular error
2148 62. When Exim is trying to find all the local interfaces, it used to panic and
2149 die if the ioctl to get the interface flags failed. However, it seems that
2150 on at least one OS (Solaris 9) it is possible to have an interface that is
2151 included in the list of interfaces, but for which you get a failure error
2152 for this call. This happens when the interface is not "plumbed" into a
2153 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2154 failure of the "get flags" call assumes that the interface is down.
2156 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2157 makes life easier for people who are doing arithmetic on fields extracted
2158 from dates, where you often get leading zeros that should not be
2159 interpreted as octal.
2161 64. Added qualify_domain to the redirect router, to override the global
2164 65. If a pathologically long header line contained very many addresses (the
2165 report of this problem mentioned 10 000) and each of them was rewritten,
2166 Exim could use up a very large amount of memory. (It kept on making new
2167 copies of the header line as it rewrote, and never released the old ones.)
2168 At the expense of a bit more processing, the header rewriting function has
2169 been changed so that it no longer eats memory in this way.
2171 66. The generation of the Received: header has been moved from the time that a
2172 message starts to be received, to the time that it finishes. The timestamp
2173 in the Received: header should now be very close to that of the <= log
2174 line. There are two side-effects of this change:
2176 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2177 logged header lines no longer include the local Received: line, because
2178 it has not yet been created. The same applies to a copy of the message
2179 that is returned to a non-SMTP sender when a message is rejected.
2181 (b) When a filter file is tested using -bf, no additional Received: header
2182 is added to the test message. After some thought, I decided that this
2185 This change does not affect the value of $received_for. It is still set
2186 after address rewriting, but before local_scan() is called.
2188 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2190 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2191 gave an unhelpful panic error message, and a defer error. I have managed to
2192 change this behaviour so that it now rejects any supplied certificate,
2193 which seems right, as the list of acceptable certificates is empty.
2195 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2196 gave an unhelpful defer error. I have not managed to make this reject any
2197 supplied certificates, but the error message it gives is "no certificate
2198 supplied", which is not helpful.
2200 70. exigrep's output now also includes lines that are not associated with any
2201 message, but which match the given pattern. Implemented by a patch from
2202 Martin Sluka, which also tidied up the Perl a bit.
2204 71. Recipient callout verification, like sender verification, was using <> in
2205 the MAIL FROM command. This isn't really the right thing, since the actual
2206 sender may affect whether the remote host accepts the recipient or not. I
2207 have changed it to use the actual sender in the callout; this means that
2208 the cache record is now keyed on a recipient/sender pair, not just the
2209 recipient address. There doesn't seem to be a real danger of callout loops,
2210 since a callout by the remote host to check the sender would use <>.
2211 [SEE ABOVE: changed after hitting problems.]
2213 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2214 temporary errors. However, in the case of such a code being given after
2215 the end of a data transmission (i.e. after ".") Exim was failing to write
2216 a retry record for the message. (Yes, there was some broken host that was
2217 actually sending 8xx at this point.)
2219 73. An unknown lookup type in a host list could cause Exim to panic-die when
2220 the list was checked. (An example that provoked this was putting <; in the
2221 middle of a list instead of at the start.) If this happened during a DATA
2222 ACL check, a -D file could be left lying around. This kind of configuration
2223 error no longer causes Exim to die; instead it causes a defer errror. The
2224 incident is still logged to the main and panic logs.
2226 74. Buglet left over from Exim 3 conversion. The message "too many messages
2227 in one connection" was written to the rejectlog but not the mainlog, except
2228 when address rewriting (yes!) was being logged.
2230 75. Added write_rejectlog option.
2232 76. When a system filter was run not as root (that is, when system_filter_user
2233 was set), the values of the $n variables were not being returned to the
2234 main process; thus, they were not subsequently available in the $sn
2237 77. Added +return_path_on_delivery log selector.
2239 78. A connection timeout was being treated differently from recipients deferred
2240 when testing hosts_max_try with a message that was older than the host's
2241 retry timeout. (The host should not be counted, thus allowing all hosts to
2242 be tried at least once before bouncing.) This may have been the cause of an
2243 occasionally reported bug whereby a message would remain on the queue
2244 longer than the retry timeout, but would be bounced if a delivery was
2245 forced. I say "may" because I never totally pinned down the problem;
2246 setting up timeout/retry tests is difficult. See also the next item.
2248 79. The ultimate address timeout was not being applied to errors that involved
2249 a combination of host plus message (for example, a timeout on a MAIL
2250 command). When an address resolved to a number of possible hosts, and they
2251 were not all tried for each delivery (e.g. because of hosts_max_try), a
2252 message could remain on the queue longer than the retry timeout.
2254 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2257 81. Fixed an obscure SMTP outgoing bug which required at least the following
2258 conditions: (a) there was another message waiting for the same server;
2259 (b) the server returned 5xx to all RCPT commands in the first message so
2260 that the message was not completed; (c) the server dropped the connection
2261 or gave a negative response to the RSET that Exim sends to abort the
2262 transaction. The observed case was a dropped connection after DATA that had
2263 been sent in pipelining mode. That is, the server had advertised PIPELINING
2264 but was not implementing it correctly. The effect of the bug was incorrect
2265 behaviour, such as trying another host, and this could lead to a crash.
2271 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2272 and daemon.c were passed as pointers to ints; they should have been
2273 pointers to socklen_t variables (which are typically unsigned ints).
2275 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2278 3. Fixed a really odd bug that affected only the testing scheme; patching a
2279 certain fixed string in the binary changed the value of another string that
2280 happened to be identical to the end of the original first string.
2282 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2283 name", it returns that address as the IP address. On some operating
2284 systems (e.g. Solaris), it also passes back the IP address string as the
2285 "host name". However, on others (e.g. Linux), it passes back an empty
2286 string. Exim wasn't checking for this, and was changing the host name to an
2287 empty string, assuming it had been canonicized.
2289 5. Although rare, it is permitted to have more than one PTR record for a given
2290 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2291 all the names associated with an address, because they do in Solaris.
2292 However, it seems that they do not in Linux for data that comes from the
2293 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2294 I found this out when I moved to a new Linux workstation and tried to run
2295 the Exim test suite.
2297 To get round this problem I have changed the code so that it now does its
2298 own call to the DNS to look up PTR records when searching for a host name.
2299 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2300 addresses that are only in /etc/hosts are still found.
2302 This behaviour is, however, controlled by an option called host_lookup_
2303 order, which defaults to "bydns:byaddr". If people want to use the other
2304 order, or indeed, just use one or the other means of lookup, they can
2305 specify it in this variable.
2307 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2308 some operating systems, this comes back from gethostbyaddr() as an empty
2309 string, and this is what Exim used to test for. However, it seems that in
2310 other systems, "." is yielded. Exim now tests for this case too.
2312 7. The values of check_spool_space and check_log_space are now held internally
2313 as a number of kilobytes instead of an absolute number of bytes. If a
2314 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2315 kilobyte. This means that much larger values can be stored.
2317 8. Exim monitor: an attempt to get the action menu when not actually pointing
2318 at a message produces an empty menu entitled "No message selected". This
2319 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2320 no entries in it ("Shell widget menu has zero width and/or height"). So I
2321 have added a single, blank menu entry in this case.
2323 9. Added ${quote_local_part.
2325 10. MIME decoding is now applied to the contents of Subject: header lines when
2328 11. Now that a reference to $sender_host_address automatically causes a reverse
2329 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2330 host lookup before query-style lookups in lists that might use this
2331 variable. This has therefore been abolished, and the "net-" prefix is no
2332 longer necessary for query-style lookups.
2334 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2335 have been a typo for LFLAGS, so it has been changed.
2337 13. The install script calls Exim with "-C /dev/null" in order to find the
2338 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2339 to be output. Howeve, since Exim outputs its version number before the
2340 error, it didn't break the script. It just looked ugly. I fixed this by
2341 always allowing "-C /dev/null" if the caller is root.
2343 14. Ignore overlarge ACL variable number when reading spool file - insurance
2344 against a later release with more variables having written the file.
2346 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2347 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2348 this, as well as the form without the "IPv6" on the front.
2350 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2351 OS/Makefile-Darwin file.
2353 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2354 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2356 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2357 variable. However, some OS use ints for inodes. I've added cast to long int
2358 to get rid of the compiler warning.
2360 19. I had forgotten to lock out "/../" in configuration file names when
2361 ALT_CONFIG_PREFIX was set.
2363 20. Routers used for verification do not need to specify transports. However,
2364 if such a router generated a host list, and callout was configured, Exim
2365 crashed, because it could not find a port number from the (non-existent)
2366 transport. It now assumes port 25 in this circumstance.
2368 21. Added the -t option to exigrep.
2370 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2371 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2372 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2373 them is set, LOOKUP_LSEARCH is forced.
2375 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2377 24. Added sender and host information to the "rejected by local_scan()" log
2378 line; previously there was no indication of these.
2380 25. Added .include_if_exists.
2382 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2383 Linux. It turns out that not all file systems support this. Apparently some
2384 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2385 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2386 not supported on the file descriptor, is now ignored when Exim is trying to
2387 sync a directory. This applies only to Linux.
2389 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2391 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2392 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2395 29. Added support for maildirsize files from supplied patch (modified a bit).
2397 30. The use of :fail: followed by an empty string could lead Exim to respond to
2398 sender verification failures with (e.g.):
2400 550 Verification failed for <xxx>
2401 550 Sender verify failed
2403 where the first response line was missing the '-' that indicates it is not
2404 the final line of the response.
2406 31. The loop for finding the name of the user that called Exim had a hardwired
2407 limit of 10; it now uses the value of finduser_retries, which is used for
2408 all other user lookups.
2410 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2413 33. Exim was neglecting to zero errno before one call of strtol() when
2414 expanding a string and expecting an integer value. On some systems this
2415 resulted in spurious "integer overflow" errors. Also, it was casting the
2416 result into an int without checking.
2418 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2419 did not work. The code looks as if it has *never* worked, though it appears
2420 to have been documented since at least releast 1.62. I have made it work.
2422 35. The "timeout_DNS" error in retry rules, also documented since at least
2423 1.62, also never worked. As it isn't clear exactly what this means, and
2424 clearly it isn't a major issue, I have abolished the feature by treating it
2425 as "timeout", and writing a warning to the main and panic logs.
2427 36. The display of retry rules for -brt wasn't always showing the error code
2430 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2431 timeout_connect_A, timeout_connect_MX.
2433 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2434 to the empty sender.
2436 39. The daemon was not analysing the content of -oX till after it had closed
2437 stderr and disconnected from the controlling terminal. This meant that any
2438 syntax errors were only noted on the panic log, and the return code from
2439 the command was 0. By re-arranging the code a little, I've made the
2440 decoding happen first, so such errors now appear on stderr, and the return
2441 code is 1. However, the actual setting up of the sockets still happens in
2442 the disconnected process, so errors there are still only recorded on the
2445 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2446 connections (as happens on some IP stacks) was logged at start up time as
2447 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2448 it from "IPv6 and IPv4", which means that two separate sockets are being
2451 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2452 says whether AF_INET or AF_INET6 was passed as an argument.
2454 42. Exiwhat output was messed up when time zones were included in log
2457 43. Exiwhat now gives more information about the daemon's listening ports,
2458 and whether -tls-on-connect was used.
2460 44. The "port" option of the smtp transport is now expanded.
2462 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2463 silently ignored. Now an error message is written to the main and panic
2466 46. There's a new ACL modifier called "logwrite" which writes to a log file
2467 as soon as it is encountered.
2469 47. Added $local_user_uid and $local_user_gid at routing time.
2471 48. Exim crashed when trying to verify a sender address that was being
2474 49. Exim was recognizing only a space character after ".include". It now also
2475 recognizes a tab character.
2477 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2478 extracting the relevant information from the specification. The man page no
2479 longer contains scrambled data for the -d option, and I've added a section
2480 at the front about calling Exim under different names.
2482 51. Added "extra_headers" argument to the "mail" command in filter files.
2484 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2487 53. Installed eximstats 1.29.
2489 54. Added transport_filter_timeout as a generic transport option.
2491 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2492 Cc: header lines. This was required by RFC 822, but it not required by RFC
2495 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2496 incoming messages that did not have them. Now it does so only if the
2497 message originates locally, that is, if there is no associated remote host
2498 address. When Resent- header lines are present, this applies to the Resent-
2499 lines rather than the non-Resent- lines.
2501 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2502 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2504 58. Messages for configuration errors now include the name of the main
2505 configuration file - useful now that there may be more than one file in a
2506 list (.included file names were always shown).
2508 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2509 for those rare installations that do not start the daemon as root or run it
2510 setuid root. I've cut out the call to initgroups() if the daemon is not
2513 60. The Exim user and group can now be bound into the binary as text strings
2514 that are looked up at the start of Exim's processing.
2516 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2518 62. Added $mailstore_basename variable.
2520 63. Installed patch to sieve.c from Michael Haardt.
2522 64. When Exim failed to open the panic log after failing to open the main log,
2523 the original message it was trying to log was written to stderr and debug
2524 output, but if they were not available (the usual case in production), it
2525 was lost. Now it is written to syslog before the two lines that record the
2526 failures to open the logs.
2528 65. Users' Exim filters run in subprocesses under the user's uid. It is
2529 possible for a "deliver" command or an alias in a "personal" command to
2530 provoke an address rewrite. If logging of address rewriting is configured,
2531 this fails because the process is not running as root or exim. There may be
2532 a better way of dealing with this, but for the moment (because 4.30 needs
2533 to be released), I have disabled address rewrite logging when running a
2534 filter in a non-root, non-exim process.
2540 1. The buildconfig auxiliary program wasn't quoting the value set for
2541 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2542 not defined. This bug was masked in 4.22 by the effect that was fixed in
2545 2. Some messages that were rejected after a message id was allocated were
2546 shown as "incomplete" by exigrep. It no longer does this for messages that
2547 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2549 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2550 have allow_domain_literals set, the ID did not get logged in the <= line.
2551 Domain literals are now always recognized in Message-ID: header lines.
2553 4. The first argument for a ${extract expansion item is the key name or field
2554 number. Leading and trailing spaces in this item were not being ignored,
2555 causing some misleading effects.
2557 5. When deliver_drop_privilege was set, single queue runner processes started
2558 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2559 same command in the process it spins off) were not dropping privilege.
2561 6. When the daemon running as "exim" started a queue runner, it always
2562 re-executed Exim in the spun-off process. This is a waste of effort when
2563 deliver_drop_privilege is set. The new process now just calls the
2564 queue-runner function directly.
2570 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2573 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2574 ignored. Though the use of -oP was forcing the writing of a pid file, it
2575 was always written to the default place.
2577 3. If the message "no IP address found for host xxxx" is generated during
2578 incoming verification, it is now followed by identification of the incoming
2579 connection (so you can more easily find what provoked it).
2581 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2583 5. Added some features to "harden" Exim a bit more against certain attacks:
2585 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2586 be put in Local/Makefile. This is like the never_users runtime option,
2587 but it cannot be overridden. The default setting is "root".
2589 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2590 prefix string with which any file named in a -C command line option
2593 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2594 is retained for -C and -D only if the caller of Exim is root. Without
2595 it, the exim user may also use -C and -D and retain privilege.
2597 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2598 command line option is disabled.
2600 6. Macro names set by the -D option must start with an upper case letter, just
2601 like macro names defined in the configuration file.
2603 7. Added "dereference=" facility to LDAP.
2605 8. Two instances of the typo "uknown" in the source files are fixed.
2607 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2608 the Configure-Makefile script screwed up while processing it.
2610 10. Incorporated PCRE 4.4.
2612 11. The SMTP synchronization check was not operating right at the start of an
2613 SMTP session. For example, it could not catch a HELO sent before the client
2614 waited for the greeting. There is now a check for outstanding input at the
2615 point when the greeting is written. Because of the duplex, asynchronous
2616 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2617 way, but not yet received, when the check is performed.
2619 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2620 on TCP/IP sockets, because this apparently causes some broken clients to
2623 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2624 unchanged) from the Cygwin maintainer.
2626 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2627 was supported instead of testing for mbx. Effectively a typo.
2629 15. The spa authenticator server code was not checking that the input it
2630 received was valid base64.
2632 16. The debug output line for the "set" modifier in ACLs was not showing the
2633 name of the variable that was being set.
2635 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2637 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2638 lookup on its own. Something else was needed to trigger the lookup. For
2639 example, a match in host_lookup or the need for a host name in a host list.
2640 Now, if $sender_host_name is referenced and the host name has not yet been
2641 looked up, a lookup is performed. If the lookup fails, the variable remains
2642 empty, and $host_lookup_failed is set to "1".
2644 19. Added "eqi" as a case-independent comparison operator.
2646 20. The saslauthd authentication condition could segfault if neither service
2647 nor realm was specified.
2649 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2650 error message that was logged was misleading, and incoming SMTP
2651 connections were dropped. The message is now more accurate, and temporary
2652 errors are given to SMTP connections.
2654 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2655 (except RSET) with a 421 error, until QUIT is received. However, it was
2656 failing to send a response to QUIT.
2658 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2659 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2660 ACL is now run after the helo_try_verify_hosts code.
2662 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2663 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2664 case-independent, so other case variants are also recognized). Apparently
2665 some systems use these upper case variants.
2667 25. If more than two messages were waiting for the same host, and a transport
2668 filter was specified for the transport, Exim sent two messages over the
2669 same TCP/IP connection, and then failed with "socket operation on non-
2670 socket" when it tried to send the third.
2672 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2674 27. The extern definition of crypt16() in expand.c was not being excluded when
2675 the OS had its own crypt16() function.
2677 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2678 as a preferred synonym for return_size_limit, both as an option and as an
2681 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2683 30. Changed the default configuration ACL to relax the local part checking rule
2684 for addresses that are not in any local domains. For these addresses,
2685 slashes and pipe symbols are allowed within local parts, but the sequence
2686 /../ is explicitly forbidden.
2688 31. SPA server authentication was not clearing the challenge buffer before
2691 32. log_message in a "warn" ACL statement was writing to the reject log as
2692 well as to the main log, which contradicts the documentation and doesn't
2693 seem right (because no rejection is happening). So I have stopped it.
2695 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2696 However, I am unable to do any testing of this.
2698 34. Fixed an infelicity in the appendfile transport. When checking directories
2699 for a mailbox, to see if any needed to be created, it was accidentally
2700 using path names with one or more superfluous leading slashes; tracing
2701 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2703 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2704 contents are added to the log line that is written for every discarded
2705 recipient. (Previously a log_message setting was ignored.)
2707 36. The ${quote: operator now quotes the string if it is empty.
2709 37. The install script runs exim in order to find its version number. If for
2710 some reason other than non-existence or emptiness, which it checks, it
2711 could not run './exim', it was installing it with an empty version number,
2712 i.e. as "exim-". This error state is now caught, and the installation is
2715 38. An argument was missing from the function that creates an error message
2716 when Exim fails to connect to the socket for saslauthd authentication.
2717 This could cause Exim to crash, or give a corrupted message.
2719 39. Added isip, isip4, and isip6 to ${if conditions.
2721 40. The ACL variables $acl_xx are now saved with the message, and can be
2722 accessed later in routers, transports, and filters.
2724 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2725 strings in the file are not string-expanded.
2727 42. If a MAIL command specified a SIZE value that was too large to fit into an
2728 int variable, the check against message_size_limit failed. Such values are
2729 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2730 day this will have to be increased, but I don't think I want to be around
2731 when emails are that large.
2738 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2739 iconv() is not standard in FreeBSD.
2741 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2742 IPv6 enabled. The observed symptom was a segmentation fault on return from
2743 the function os_common_find_running_interfaces() in src/os.c.
2745 3. In the check_special_case() function in daemon.c I had used "errno" as an
2746 argument name, which causes warnings on some systems. This was basically a
2747 typo, since it was named "eno" in the comments!
2749 4. The code that waits for the clock to tick (at a resolution of some fraction
2750 of a second) so as to ensure message-id uniqueness was always waiting for
2751 at least one whole tick, when it could have waited for less. [This is
2752 almost certainly not relevant at current processor speeds, where it is
2753 unlikely to ever wait at all. But we try to future-proof.]
2755 5. The function that sleeps for a time interval that includes fractions of a
2756 second contained a race. It did not block SIGALRM between setting the
2757 timer, and suspending (a couple of lines later). If the interval was short
2758 and the sigsuspend() was delayed until after it had expired, the suspension
2759 never ended. On busy systems this could lead to processes getting stuck for
2762 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2763 process, before it forks any delivery processes. The open lookup caching
2764 mechanism meant that the open file or database connection was passed into
2765 the delivery process. The problem was that delivery processes always tidy
2766 up cached lookup data. This could cause a problem for the next delivery
2767 process started by the queue runner, because the external queue runner
2768 process does not know about the closure. So the next delivery process
2769 still has data in the lookup cache. In the case of a file lookup, there was
2770 no problem because closing a file descriptor in a subprocess doesn't affect
2771 the parent. However, if the lookup was caching a connection to a database,
2772 the connection was closed, and the second delivery process was likely to
2773 see errors such as "PGSQL: query failed: server closed the connection
2774 unexpectedly". The problem has been fixed by closing all cached lookups
2775 in a queue runner before running a delivery process.
2777 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2778 seem to have the "const" qualifier which it has on other OS. I've
2781 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2782 *of the same type* (client or server) with the same public name that an
2783 error should be diagnosed.
2785 9. When Exim looked up a host name for an IP address, but failed to find the
2786 original IP address when looking up the host name (a safety check), it
2787 output the message "<ip address> does not match any IP for NULL", which was
2788 confusing, to say the least. The bug was that the host name should have
2789 appeared instead of "NULL".
2791 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2792 user that is built into the binary, and the -C or -D options is used, root
2793 privilege is dropped before the configuration file is read. In addition,
2794 logging is switched to stderr instead of the normal log files. If the
2795 configuration then re-defines the Exim user, the unprivileged environment
2796 is probably not what is expected, so Exim logs a panic warning message (but
2799 However, if deliver_drop_privilege is set, the unprivileged state may well
2800 be exactly what is intended, so the warning has been cut out in that case,
2801 and Exim is allowed to try to write to its normal log files.
2807 1. smtp_return_error_details was not giving details for temporary sender
2808 or receiver verification errors.
2810 2. Diagnose a configuration error if two authenticators have the same public
2813 3. Exim used not to create the message log file for a message until the first
2814 delivery attempt. This could be confusing when incoming messages were held
2815 for policy or load reasons. The message log file is now created at the time
2816 the message is received, and an initial "Received" line is written to it.
2818 4. The automatically generated man page for command line options had a minor
2819 bug that caused no ill effects; however, a more serious problem was that
2820 the procedure for building the man page automatically didn't always
2821 operate. Consequently, release 4.20 contains an out-of-date version. This
2822 shouldn't happen again.
2824 5. When building Exim with embedded Perl support, the script that builds the
2825 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2826 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2828 6. The freeze_tell option was not being used for messages that were frozen on
2829 arrival, either by an ACL or by local_scan().
2831 7. Added the smtp_incomplete_transaction log selector.
2833 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2834 was accepting AUTH without a new EHLO.
2836 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2837 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2840 10. Logging of TCP/IP connections (when configured) now happens in the main
2841 daemon process instead of the child process, so that the TCP/IP connection
2842 count is more accurate (but it can never be perfect).
2844 11. The use of "drop" in a nested ACL was not being handled correctly in the
2845 outer ACL. Now, if condition failure induced by the nested "drop" causes
2846 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2847 or "require"), the connection is dropped.
2849 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2850 that yield "discard" can now be used with an "accept" or a "discard" verb,
2851 but an error is generated for any others (because I can't see a useful way
2852 to define what should happen).
2854 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2855 are now processed in the same way as lines in the Exim configuration file.
2856 In particular, continuation lines are supported.
2858 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2860 15. Added -ti meaning -t -i.
2862 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2863 domains, and warn by logging if others are found.
2865 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2866 returned from GIFCONF when Exim is trying to find the list of interfaces on
2867 a host. The code in os.c has been modified to copy each ifreq to an aligned
2868 structure in all cases.
2870 Also, in some cases, the returned ifreq's were being copied to a 'struct
2871 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2872 means the last couple of bytes of an IPv6 address could be chopped if the
2873 ifreq contained only a normal sockaddr (14 bytes storage).
2875 18. Named domain lists were not supported in the hosts_treat_as_local option.
2876 An entry such as +xxxx was not recognized, and was treated as a literal
2879 19. Ensure that header lines added by a DATA ACL are included in the reject log
2880 if the ACL subsequently rejects the message.
2882 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2883 MD5 (which is deprecated).
2885 21. When testing a filter file using -bf, Exim was writing a message when it
2886 took the sender from a "From " line in the message, but it was not doing so
2887 when it took $return_path from a Return-Path: header line. It now does.
2889 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2890 with a valid header line field name (a series of printing characters
2891 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2893 23. Changed "disc" in the source to "disk" to conform to the documentation and
2894 the book and for uniformity.
2896 24. Ignore Sendmail's -Ooption=value command line item.
2898 25. When execve() failed while trying to run a command in a pipe transport,
2899 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2900 could be confused with a return value of 69 from the command itself. This
2901 has been changed to 127, the value the shell returns if it is asked to run
2902 a non-existent command. The wording for the related log line suggests a
2903 non-existent command as the problem.
2905 26. If received_header_text expands to an empty string, do not add a Received:
2906 header line to the message. (Well, it adds a token one on the spool, but
2907 marks it "old" so that it doesn't get used or transmitted.)
2909 27. Installed eximstats 1.28 (addition of -nt option).
2911 28. There was no check for failure on the call to getsockname() in the daemon
2912 code. This can fail if there is a shortage of resources on the system, with
2913 ENOMEM, for example. A temporary error is now given on failure.
2915 29. Contrary to the C standard, it seems that in some environments, the
2916 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2917 program. Exim now does this explicitly; it affects the formatting of
2918 timestamps using strftime().
2920 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2921 complaints. I've now initialized all the variables, to avoid this.
2923 32. Header lines added by a system filter were not being "seen" during
2924 transport-time rewrites.
2926 33. The info_callback() function passed to OpenSSL is set up with type void
2927 (*)(SSL *, int, int), as described somewhere. However, when calling the
2928 function (actually a macro) that sets it up, the type void(*)() is
2929 expected. I've put in a cast to prevent warnings from picky compilers.
2931 34. If a DNS black list lookup found a CNAME record, but there were no A
2932 records associated with the domain it pointed at, Exim crashed.
2934 35. If a DNS black list lookup returned more than one A record, Exim ignored
2935 all but the first. It now scans all returned addresses if a particular IP
2936 value is being sought. In this situation, the contents of the
2937 $dnslist_value variable are a list of all the addresses, separated by a
2940 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2941 to accept a host name and all its aliases if the forward lookup for any of
2942 them yielded the IP address of the incoming connection. Now it accepts only
2943 those names whose forward lookup yields the correct IP address. Any other
2944 names are discarded. This closes a loophole whereby a rogue DNS
2945 administrator could create reverse DNS records to break through a
2946 wildcarded host restriction in an ACL.
2948 37. If a user filter or a system filter that ran in a subprocess used any of
2949 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2950 the wrong values were passed to the pipe command ($thisaddress had the
2951 value of $0, $0 had the value of $1, etc). This bug was introduced by
2952 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2954 38. Improved the line breaking for long SMTP error messages from ACLs.
2955 Previously, if there was no break point between 40 and 75 characters, Exim
2956 left the rest of the message alone. Two changes have been made: (a) I've
2957 reduced the minimum length to 35 characters; (b) if it can't find a break
2958 point between 35 and 75 characters, it looks ahead and uses the first one
2959 that it finds. This may give the occasional overlong line, but at least the
2960 remaining text gets split now.
2962 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2963 file descriptors might be low, and that setting 1000 would always raise it.
2964 It turns out that in some environments, the limit is already over 1000 and
2965 that lowering it causes trouble. So now Exim takes care not to decrease it.
2967 40. When delivering a message, the value of $return_path is set to $sender_
2968 address at the start of routing (routers may change the value). By an
2969 oversight, this default was not being set up when an address was tested by
2970 -bt or -bv, which affected the outcome if any router or filter referred to
2973 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2974 log only when "message" or "log_message" are set. However, if one of the
2975 conditions was an address verification, or a call to a nested ACL, the
2976 messages generated by the underlying test were being passed through. This
2977 no longer happens. The underlying message is available in $acl_verify_
2978 message for both "message" and "log_message" expansions, so it can be
2979 passed through if needed.
2981 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2982 new expansion $bh_ to give the encoded byte string without charset
2983 translation. Translation happens only if iconv() is available; HAVE_ICONV
2984 indicates this at build time. HEADERS_CHARSET gives the charset to
2985 translate to; headers_charset can change it in the configuration, and
2986 "headers charset" can change it in an individual filter file.
2988 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
2989 that creates RFC 2047 encoded "words" labels them as that charset instead
2990 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2991 expansion operator; (ii) when Exim creates a From: line for a local
2992 message; (iii) when a header line is rewritten to include a "phrase" part.
2994 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2995 buggy, causing it to skip the first lines of messages whose message ID
2996 ended in 'D'. This would not have bitten before Exim release 4.14, because
2997 message IDs were unlikely to end in 'D' before then. The effect was to have
2998 incorrect size information for certain domains.
3000 45. #include "config.h" was missing at the start of the crypt16.c module. This
3001 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3004 46. If there was a timeout during a "random" callout check, Exim treated it as
3005 a failure of the random address, and carried on sending RSET and the real
3006 address. If the delay was just some slowness somewhere, the response to the
3007 original RCPT would be taken as a response to RSET and so on, causing
3008 mayhem of various kinds.
3010 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3011 when I implemented it. It didn't allow for the fact that some option values
3012 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3013 the right test for positive values.
3015 48. Domain names in DNS records are case-independent. Exim always looks them up
3016 in lower case. Some resolvers return domain names in exactly the case they
3017 appear in the zone file, that is, they may contain uppercase letters. Not
3018 all resolvers do this - some return always lower case. Exim was treating a
3019 change of case by a resolver as a change of domain, similar to a widening
3020 of a domain abbreviation. This triggered its re-routing code and so it was
3021 trying to route what was effectively the same domain again. This normally
3022 caused routing to fail (because the router wouldn't handle the domain
3023 twice). Now Exim checks for this case specially, and just changes the
3024 casing of the domain that it ultimately uses when it transmits the message
3027 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3030 50. If a filter generated a file delivery with a non-absolute name (possible if
3031 no home directory exists for the router), the forbid_file option was not
3034 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3035 the existing equality matching.
3037 52. Exim was using ints instead of ino_t variables in some places where it was
3038 dealing with inode numbers.
3040 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3041 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3042 called TMPDIR, and if it finds it is different, it changes its value.
3044 54. The smtp_printf() function is now made available to local_scan() so
3045 additional output lines can be written before returning. There is also an
3046 smtp_fflush() function to enable the detection of a dropped connection.
3047 The variables smtp_input and smtp_batched_input are exported to
3050 55. Changed the default runtime configuration: the message "Unknown user"
3051 has been removed from the ACL, and instead placed on the localuser router,
3052 using the cannot_route_message feature. This means that any verification
3053 failures that generate their own messages won't get overridden. Similarly,
3054 the "Unrouteable address" message that was in the ACL for unverifiable
3055 relay addresses has also been removed.
3057 56. Added hosts_avoid_esmtp to the smtp transport.
3059 57. The exicyclog script was not checking for the esoteric option
3060 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3061 will work only if exicyclog is run under the appropriate euid.
3063 58. Following a discussion on the list, the rules by which Exim recognises line
3064 endings on incoming messages have been changed. The -dropcr and drop_cr
3065 options are now no-ops, retained only for backwards compatibility. The
3066 following line terminators are recognized: LF CRLF CR. However, special
3067 processing applies to CR:
3069 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3070 nor a local message in the state where . is a terminator.
3072 (ii) If a bare CR is encountered in a header line, an extra space is added
3073 after the line terminator so as not to end the header. The reasoning
3074 behind this is that bare CRs in header lines are most likely either
3075 to be mistakes, or people trying to play silly games.
3077 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3078 was being incorrectly given as 18 bytes larger than it should have been.
3079 This is a VOB (very old bug).
3081 60. This may never have affected anything current, but just in case it has:
3082 When the local host is found other than at the start of a list of hosts,
3083 the local host, those with the same MX, and any that follow, are discarded.
3084 When the list in question was part of a longer list of hosts, the following
3085 hosts (not currently being processed) were also being discarded. This no
3086 longer happens. I'm not sure if this situation could ever has previously
3089 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3092 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3093 header. This is something that is recommended in a new Internet Draft, and
3094 is something that is documented as being done by Sendmail. There are two
3095 possible values. For messages generated by the autoreply transport, Exim
3098 Auto-Submitted: auto-replied
3100 whereas for all other generated messages (e.g. bounces) it adds
3102 Auto-Submitted: auto-generated
3104 63. The "personal" condition in filters now includes a test for the
3105 Auto-Submitted: header. If it contains the string "auto-" the message it
3106 not considered personal.
3108 64. Added rcpt_include_affixes as a generic transport option.
3110 65. Added queue_only_override (default true).
3112 66. Added the syslog_duplication option.
3114 67. If what should have been the first header line of a message consisted of
3115 a space followed by a colon, Exim was mis-interpreting it as a header line.
3116 It isn't of course - it is syntactically invalid and should therefore be
3117 treated as the start of the message body. The misbehaviour could have
3118 caused a number of strange effects, including loss of data in subsequent
3119 header lines, and spool format errors.
3121 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3122 client host had authenticated. This control can now be exercised by an ACL
3123 for more flexibility.
3125 69. By default, callouts do not happen when testing with -bh. There is now a
3126 variant, -bhc, which does actually run the callout code, including
3127 consulting and updating the callout cache.
3129 70. Added support for saslauthd authentication, courtesy of Alexander
3132 71. If statvfs() failed on the spool or log directories while checking their
3133 size for availability, Exim confusingly gave the error "space shortage".
3134 Furthermore, in debugging mode it crashed with a floating point exception.
3135 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3136 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3137 serious problem, Exim now writes to the main and panic logs when this
3138 happens, with details of the failure. It then refuses to accept the
3139 incoming message, giving the message "spool directory problem" or "log
3140 directory problem" with a 421 code for SMTP messages.
3142 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3143 0, 1, and 2 exist, because some OS complain for execs without them (see
3144 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3145 descriptors. However, the code omitted to check that the open succeeded,
3146 causing mysterious errors if for some reason the permissions on /dev/null
3147 got screwed. Now Exim writes a message to the main and panic logs, and
3148 bombs out if it can't open /dev/null.
3150 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3151 interact so that it is all more flexible. It is supposed to remain
3152 backwards compatible. Also added extra_local_interfaces.
3154 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3155 to bomb out with an assertion failure - to the client this appears as a
3156 connection drop. This problem occurs in the part of the code that was taken
3157 from the Samba project. Fortunately, the assertion is in a very simple
3158 function, so I have fixed this by reproducing the function inline in the
3159 one place where it is called, and arranging for authentication to fail
3160 instead of killing the process with assert().
3162 75. The SPA client code was not working when the server requested OEM rather
3163 than Unicode encoding.
3165 76. Added code to make require_files with a specific uid setting more usable in
3166 the case where statting the file as root fails - usually a non-root-mounted
3167 NFS file system. When this happens and the failure is EACCES, Exim now
3168 forks a subprocess and does the per-uid checking as the relevant uid.
3170 77. Added process_log_path.
3172 78. If log_file_path was not explicitly set, a setting of check_log_space or
3173 check_log_inodes was ignored.
3175 79. If a space check for the spool or log partitions fails, the incident is now
3176 logged. Of course, in the latter case the data may get lost...
3178 80. Added the %p formatting code to string_format() so that it can be used to
3179 print addresses in debug_print(). Adjusted all the address printing in the
3180 debugging in store.c to use %p rather than %d.
3182 81. There was a concern that a line of code in smtp_in.c could overflow a
3183 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3184 initially expressed, the concern was not well-founded, because trailing
3185 spaces are removed early. However, if the trailing spaces were followed by
3186 a NULL, they did not get removed, so the overflow was possible. Two fixes
3189 (a) I re-wrote the offending code in a cleaner fashion.
3190 (b) If an incoming SMTP command contains a NULL character, it is rejected
3193 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3194 runs initgroups(), so that if the Exim user is in any additional groups,
3195 they will be used during message reception.
3201 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob