1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.190 2005/08/02 09:01:44 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
42 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
44 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
45 command itself ended in error, the underlying message about the transport
46 filter timeout was being overwritten with the pipe command error. Now the
47 underlying error message should be appended to the second error message.
49 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
51 PH/02 Reduce the amount of output that "make" produces by default. Full output
52 can still be requested.
54 PH/03 The warning log line about a condition test deferring for a "warn" verb
55 was being output only once per connection, rather than after each
56 occurrence (because it was using the same function as for successful
57 "warn" verbs). This seems wrong, so I have changed it.
59 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
60 it should not have, which might have caused a crash in the right
61 circumstances, but probably never did.
63 PH/04 Installed a modified version of Tony Finch's patch to make submission
64 mode fix the return path as well as the Sender: header line, and to
65 add a /name= option so that you can make the user's friendly name appear
68 TF/03 Added the control = fakedefer ACL modifier.
70 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
71 Mark Lowes for thorough testing.
73 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
75 TK/03 Merged latest SRS patch from Miles Wilton.
77 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
78 with the definition in sysexits.h (which is #included earlier).
79 Fortunately, Exim does not actually use EX_OK. The code used to try to
80 preserve the sysexits.h value, by assumimg that macro definitions were
81 scanned for macro replacements. I have been disabused of this notion,
82 so now the code just undefines EX_OK before #including unistd.h.
84 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
85 in the smtp transport. When a block could not be written in a single
86 write() function, the timeout was being re-applied to each part-write.
87 This seems wrong - if the receiver was accepting one byte at a time it
88 would take for ever. The timeout is now adjusted when this happens. It
89 doesn't have to be particularly precise.
91 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
92 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
94 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
95 without a domain if the check to <postmaster@domain> fails.
97 SC/01 Eximstats: added -xls and the ability to specify output files
98 (patch written by Frank Heydlauf).
100 SC/02 Eximstats: use FileHandles for outputing results.
102 SC/03 Eximstats: allow any combination of xls, txt, and html output.
104 SC/04 Eximstats: fixed display of large numbers with -nvr option
106 SC/05 Eximstats: fixed merging of reports with empty tables.
108 SC/06 Eximstats: added the -include_original_destination flag
110 SC/07 Eximstats: removed tabs and trailing whitespace.
112 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
114 TK/06 MBOX spool code: Add real "From " MBOX separator line
115 so the .eml file is really in mbox format (even though
116 most programs do not really care). Patch from Alex Miller.
118 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
119 The latter is generated from $received_to and is only set if the
120 message has one envelope recipient. SA can use these headers,
121 obviously out-of-the-box. Patch from Alex Miller.
123 PH/08 The ${def test on a variable was returning false if the variable's
124 value was "0", contrary to what the specification has always said!
125 The result should be true unless the variable is empty.
127 PH/09 The syntax error of a character other than { following "${if
128 def:variable_name" (after optional whitespace) was not being diagnosed.
129 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
130 accidental colon was present, for example, could give incorrect results.
132 PH/10 Tidied the code in a number of places where the st_size field of a stat()
133 result is used (not including appendfile, where other changes are about
136 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
137 This involved changing a lot of size variables from int to off_t. It
138 should work with maildirs and everything.
140 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
141 spamd dying while we are connected to it.
143 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
146 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
147 function for BDB 4.3.
149 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
151 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
152 the sources for PCRE have been re-arranged and more files are now
155 PH/15 The code I had for printing potentially long long variables in PH/11
156 above was not the best (it lost precision). The length of off_t variables
157 is now inspected at build time, and an appropriate printing format (%ld
158 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
159 to be "long long int" or "long int". This is needed for the internal
160 formatting function string_vformat().
162 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
163 the configuration file to be ":syslog", then the script "guesses" where
164 the logs files are, rather than using the compiled in default. In our
165 case the guess is not the same as the compiled default, so the script
166 suddenly stopped working when I started to use syslog. The patch checks
167 to see if log_file_path is "". If so, it attempts to read it from exim
168 with no configuration file to get the compiled in version, before it
169 falls back to the previous guessing code."
171 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
172 implementing BATV in an Exim configuration. See NewStuff for the gory
175 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
176 Makefile that are specific to HP-UX.
178 PH/18 If the "use_postmaster" option was set for a recipient callout together
179 with the "random" option, the postmaster address was used as the MAIL
180 FROM address for the random test, but not for the subsequent recipient
181 test. It is now used for both.
183 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
184 patch removes a few documentation additions to RFC 3028, because the
185 latest draft now contains them. It adds the new en;ascii-case comparator
186 and a new error check for 8bit text in MIME parts. Comparator and
187 require names are now matched exactly. I enabled the subaddress
188 extension, but it is not well tested yet (read: it works for me)."
190 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
191 rework some of the code of TK/09 above to avoid the hardwired use of
192 "%lld" and "long long". Replaced the call to snprintf() with a call to
195 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
196 records point to non-existent hosts", "retry timeout exceeded", and
197 "retry time not reached for any host after a long failure period".
199 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
200 experimental DomainKeys support:
202 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
203 (2) On an error such as an illegally used "control", the wrong name for
204 the control was given.
206 These problems did NOT occur unless DomainKeys support was compiled.
208 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
210 PH/24 Added ${if match_ip condition.
212 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
213 databases so that it will be absolutely obvious if a crash occurs in the
214 DB library. This is a regular occurrence (often caused by mis-matched
217 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
218 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
219 picked up on a user's system that detects such things. There doesn't seem
220 to be a gcc warning option for this - only an attribute that has to be
221 put on the function's prototype. It seems that in Fedora Core 4 they have
222 set this on a number of new functions. No doubt there will be more in due
225 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
226 specify a transport. However, if an address that was verified by such a
227 router was the subject of a callout, Exim crashed because it tried to
228 read the rcpt_include_affixes from the non-existent transport. Now it
229 just assumes that the setting of that option is false. This bug was
230 introduced by 4.51/PH/31.
232 PH/28 Changed -d+all to exclude +memory, because that information is very
233 rarely of interest, but it makes the output a lot bigger. People tend to
234 do -d+all out of habit.
236 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
237 code in os-type was giving problems when libc.so lives in lib64, like on
240 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
241 aren't the modern standard, and it seems that some systems' include files
242 don't always have them. Exim was already checking for some of the newer
243 ones like T_AAAA, and defining it itself. I've added checks for all the
244 record types that Exim uses.
246 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
247 not automatically generating a new one, as it is supposed to. This
248 prevented TLS from working. If the file did exist, but contained invalid
249 data, a new version was generated, as expected. It was only the case of a
250 non-existent file that was broken.
252 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
253 with a change in libdomainkeys > 0.64.
255 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
256 from DNS. If the selector record carries the flag, it now has
257 precedence over the domain-wide flag.
259 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
261 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
262 the use of a port name that isn't defined in /etc/services) occurred, the
263 message was deferred as in a normal delivery, and thus remained on the
264 spool, instead of being failed because of the mua_wrapper setting. This
265 is now fixed, and I tidied up some of the mua_wrapper messages at the
268 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
269 the messages in a hash of arrays rather than using individual hashes.
270 This is a bit cleaner and results in dramatic memory savings, albeit
271 at a slight CPU cost.
273 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
274 as requested by Marc Sherman.
276 SC/10 Eximstats: added histograms for user specified patterns as requested
279 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
281 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
282 fopen() in the content-scanning modules that did not already have it.
288 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
289 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
291 TK/02 Fix ACL "control" statement not being available in MIME ACL.
293 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
295 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
296 to test Sieve filters that use "vacation".
298 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
299 that changes the way the GnuTLS parameters are stored in the cache file.
300 The new format can be generated externally. For backward compatibility,
301 if the data in the cache doesn't make sense, Exim assumes it has read an
302 old-format file, and it generates new data and writes a new file. This
303 means that you can't go back to an older release without removing the
306 PH/03 A redirect router that has both "unseen" and "one_time" set does not
307 work if there are any delivery delays because "one_time" forces the
308 parent to be marked "delivered", so its unseen clone is never tried
309 again. For this reason, Exim now forbids the simultaneous setting of
312 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
313 redirected to themselves ("homonym" addresses). Read the long ChangeLog
314 entry if you want to know the details. The fix, however, neglected to
315 consider the case when local delivery batching is involved. The test for
316 "previously delivered" was not happening when checking to see if an
317 address could be batched with a previous (undelivered) one; under
318 certain circumstances this could lead to multiple deliveries to the same
321 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
322 in its include files, and this causes problems building Exim.
324 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
325 header_syntax) but Exim was just ignoring anything given after a slash.
326 In particular, this caused confusion with an attempt to use "verify =
327 reverse_host_lookup/defer_ok". An error is now given when options are
328 supplied for verify items that do not have them. (Maybe reverse_host_
329 lookup should have a defer_ok option, but that's a different point.)
331 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
332 defined by RFC 821) to 2048, because there were problems with some AUTH
333 commands, and RFC 1869 says the size should be increased for extended
334 SMTP commands that take arguments.
336 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
339 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
340 "unknown" error; now it says that the functionality isn't in the binary.
342 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
343 an address' error message when a string expansion fails (syntax or
344 whatever). Otherwise the password may appear in the log. Following change
345 PH/42 below, there is no longer a chance of it appearing in a bounce
348 PH/11 Installed exipick version 20050225.0 from John Jetmore.
350 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
351 of its addresses was ever tried. (Bugzilla bug #2.)
353 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
354 the result incorrectly in the debug output. (It correctly added a newline
355 to what was transported.)
357 TF/01 Added $received_time.
359 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
360 commented out examples of how to interface to a virus scanner and to
361 SpamAssassin. Also added commented examples of av_scanner and
362 spamd_address settings.
364 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
365 and controls are allowed in which ACLs. There were a couple of minor
366 errors. Some of the entries in the conditions table (which is a table of
367 where they are NOT allowed) were getting very unwieldy; rewrote them as a
368 negation of where the condition IS allowed.
370 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
372 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
373 header file does not have a version number, so I've had to invent a new
374 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
375 API. The code is untested by me (my Linux distribution still has 0.3.2 of
376 radiusclient), but it was contributed by a Radius user.
378 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
379 files or directories, for OpenSSL.
381 PH/19 When an Exim process that is running as root has to create an Exim log
382 file, it does so in a subprocess that runs as exim:exim so as to get the
383 ownership right at creation (otherwise, other Exim processes might see
384 the file with the wrong ownership). There was no test for failure of this
385 fork() call, which would lead to the process getting stuck as it waited
386 for a non-existent subprocess. Forks do occasionally fail when resources
387 run out. I reviewed all the other calls to fork(); they all seem to check
390 PH/20 When checking for unexpected SMTP input at connect time (before writing
391 the banner), Exim was not dealing correctly with a non-positive return
392 from the read() function. If the client had disconnected by this time,
393 the result was a log entry for a synchronization error with an empty
394 string after "input=" when read() returned zero. If read() returned -1
395 (an event I could not check), uninitialized data bytes were printed.
396 There were reports of junk text (parts of files, etc) appearing after
399 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
401 PH/22 Added support for macro redefinition, and (re)definition in between
402 driver and ACL definitions.
404 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
405 forgetting to use the resulting value; it was using the unexpanded value.
407 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
408 hadn't been configured. The fix is from Juergen Kreileder, who
409 understands it better than I do:
411 "Here's what I see happening with three configured cyrus_sasl
412 authenticators configured (plain, login, cram-md5):
414 On startup auth_cyrus_sasl_init() gets called for each of these.
415 This means three calls to sasl_listmech() without a specified mech_list.
416 => SASL tests which mechs of all available mechs actually work
417 => three warnings about OTP not working
418 => the returned list contains: plain, login, cram-md5, digest-md5, ...
420 With the patch, sasl_listmech() also gets called three times. But now
421 SASL's mech_list option is set to the server_mech specified in the the
422 authenticator. Or in other words, the answer from sasl_listmech()
423 gets limited to just the mech you're testing for (which is different
425 => the return list contains just 'plain' or 'login', 'cram-md5' or
426 nothing depending on the value of ob->server_mech.
428 I've just tested the patch: Authentication still works fine,
429 unavailable mechs specified in the exim configuration are still
430 caught, and the auth.log warnings about OTP are gone."
432 PH/25 When debugging is enabled, the contents of the command line are added
433 to the debugging output, even when log_selector=+arguments is not
436 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
437 answer is "GNU", and only if the return is "GNU/something" is the answer
440 PH/27 $acl_verify_message is now set immediately after the failure of a
441 verification in an ACL, and so is available in subsequent modifiers. In
442 particular, the message can be preserved by coding like this:
444 warn !verify = sender
445 set acl_m0 = $acl_verify_message
447 Previously, $acl_verify_message was set only while expanding "message"
448 and "log_message" when a very denied access.
450 PH/28 Modified OS/os.c-Linux with
452 -#ifndef OS_LOAD_AVERAGE
453 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
455 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
456 nomenclature these days.)
458 PH/29 Installed patch from the Sieve maintainer that adds the options
459 sieve_useraddress and sieve_subaddress to the redirect router.
461 PH/30 In these circumstances:
462 . Two addresses routed to the same list of hosts;
463 . First host does not offer TLS;
464 . First host accepts first address;
465 . First host gives temporary error to second address;
466 . Second host offers TLS and a TLS session is established;
467 . Second host accepts second address.
468 Exim incorrectly logged both deliveries with the TLS parameters (cipher
469 and peerdn, if requested) that were in fact used only for the second
472 PH/31 When doing a callout as part of verifying an address, Exim was not paying
473 attention to any local part prefix or suffix that was matched by the
474 router that accepted the address. It now behaves in the same way as it
475 does for delivery: the affixes are removed from the local part unless
476 rcpt_include_affixes is set on the transport.
478 PH/32 Add the sender address, as F=<...>, to the log line when logging a
479 timeout during the DATA phase of an incoming message.
481 PH/33 Sieve envelope tests were broken for match types other than :is. I have
482 applied a patch sanctioned by the Sieve maintainer.
484 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
485 the uid or gid is negative. A case of a negative gid caused this to be
486 noticed. The fix allows for either to be negative.
488 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
489 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
492 PH/36 The change PH/12 above was broken. Fixed it.
494 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
495 the grounds that routing the same address twice would always produce the
496 same answer. This might have been true once, but it is certainly no
497 longer true now. Routing a child address may depend on the previous
498 routing that produced that child. Some complicated redirection strategies
499 went wrong when messages had multiple recipients, and made Exim's
500 behaviour dependent on the order in which the addresses were given.
502 I have moved the duplicate checking until after the routing is complete.
503 Exim scans the addresses that are assigned to local and remote
504 transports, and removes any duplicates. This means that more work will be
505 done, as duplicates will always all be routed, but duplicates are
506 presumably rare, so I don't expect this is of any significance.
508 For deliveries to pipes, files, and autoreplies, the duplicate checking
509 still happens during the routing process, since they are not going to be
512 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
513 It corrects a timeout issue with spamd. This is Ian's comment: "The
514 background is that sometimes spamd either never reads data from a
515 connection it has accepted, or it never writes response data. The exiscan
516 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
517 blindly assumes that writes won't block so it may never time out."
519 PH/39 Allow G after quota size as well as K and M.
521 PH/40 The value set for $authenticated_id in an authenticator may not contain
522 binary zeroes or newlines because the value is written to log lines and
523 to spool files. There was no check on this. Now the value is run through
524 the string_printing() function so that such characters are converted to
525 printable escape sequences.
527 PH/41 $message_linecount is a new variable that contains the total number of
528 lines in the message. Compare $body_linecount, which is the count for the
531 PH/42 Exim no longer gives details of delivery errors for specific addresses in
532 bounce and delay warning messages, except in certain special cases, which
535 (a) An SMTP error message from a remote host;
536 (b) A message specified in a :fail: redirection;
537 (c) A message specified in a "fail" command in a system filter;
538 (d) A message specified in a FAIL return from the queryprogram router;
539 (e) A message specified by the cannot_route_message router option.
541 In these cases only, Exim does include the error details in bounce and
542 warning messages. There are also a few cases where bland messages such
543 as "unrouteable address" or "local delivery error" are given.
545 PH/43 $value is now also set for the "else" part of a ${run expansion.
547 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
548 being worked on, but at least Exim now implements the latest version to
551 PH/45 In a pipe transport, although a timeout while waiting for the pipe
552 process to complete was treated as a delivery failure, a timeout while
553 writing the message to the pipe was logged, but erroneously treated as a
554 successful delivery. Such timeouts include transport filter timeouts. For
555 consistency with the overall process timeout, these timeouts are now
556 treated as errors, giving rise to delivery failures by default. However,
557 there is now a new Boolean option for the pipe transport called
558 timeout_defer, which, if set TRUE, converts the failures into defers for
559 both kinds of timeout. A transport filter timeout is now identified in
562 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
563 systems where "make" and "gmake" are different, calling "gmake" at top
564 level broke things. I've arranged for the value of $(MAKE) to be passed
565 from the Makefile to this script so that it can call the same version of
569 A note about Exim versions 4.44 and 4.50
570 ----------------------------------------
572 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
573 changes of various kinds. As a consequence, a big documentation update was
574 needed. This delayed the release for rather longer than seemed good, especially
575 in the light of a couple of (minor) security issues. Therefore, the changes
576 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
577 release. So 4.44 and 4.50 are in effect two different branches that both start
580 I have left the 4.50 change log unchanged; it contains all the changes since
581 4.43. The change log for 4.44 is below; many of its items are identical to
582 those for 4.50. This seems to be the most sensible way to preserve the
583 historical information.
589 1. Minor wording change to the doc/README.SIEVE file.
591 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
592 computation of the current number of files was incorrect.
594 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
595 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
598 4. Give more explanation in the error message when the command for a transport
599 filter fails to execute.
601 5. There are several places where Exim runs a non-Exim command in a
602 subprocess. The SIGUSR1 signal should be disabled for these processes. This
603 was being done only for the command run by the queryprogram router. It is
604 now done for all such subprocesses. The other cases are: ${run, transport
605 filters, and the commands run by the lmtp and pipe transports.
607 6. Added CONFIGURE_GROUP build-time option.
609 7. Some older OS have a limit of 256 on the maximum number of file
610 descriptors. Exim was using setrlimit() to set 1000 as a large value
611 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
612 systems. I've change it so that if it can't get 1000, it tries for 256.
614 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
615 was an oversight, and furthermore, ever since the addition of extra
616 controls (e.g. 4.43/32), the checks on when to allow different forms of
617 "control" were broken. There should now be diagnostics for all cases when a
618 control that does not make sense is encountered.
620 9. Added the /retain_sender option to "control=submission".
622 10. $recipients is now available in the predata ACL (oversight).
624 11. Tidy the search cache before the fork to do a delivery from a message
625 received from the command line. Otherwise the child will trigger a lookup
626 failure and thereby defer the delivery if it tries to use (for example) a
627 cached ldap connection that the parent has called unbind on.
629 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
630 of $address_data from the recipient verification was clobbered by the
633 13. The value of address_data from a sender verification is now available in
634 $sender_address_data in subsequent conditions in the ACL statement.
636 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
638 15. Added a new option "connect=<time>" to callout options, to set a different
641 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
642 was its contents. (It was OK if the option was not defined at all.)
644 17. A "Completed" log line is now written for messages that are removed from
645 the spool by the -Mrm option.
647 18. New variables $sender_verify_failure and $recipient_verify_failure contain
648 information about exactly what failed.
650 19. Added -dd to debug only the daemon process.
652 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
653 handles timeouts, both on the server side and network timeouts. Renamed the
654 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
656 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
657 It is now set to "smtps".
659 22. $host_address is now set to the target address during the checking of
662 23. When checking ignore_target_hosts for an ipliteral router, no host name was
663 being passed; this would have caused $sender_host_name to have been used if
664 matching the list had actually called for a host name (not very likely,
665 since this list is usually IP addresses). A host name is now passed as
668 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
669 code that specifies a non-restarting handler (typically sigaction() in
670 modern systems) in an attempt to fix a rare and obscure crash bug.
672 25. Narrowed the window for a race in the daemon that could cause it to ignore
673 SIGCHLD signals. This is not a major problem, because they are used only to
674 wake it up if nothing else does.
676 26. A malformed maildirsize file could cause Exim to calculate negative values
677 for the mailbox size or file count. Odd effects could occur as a result.
678 The maildirsize information is now recalculated if the size or filecount
681 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
682 support for a long time. Removed HAVE_SYS_VFS_H.
684 28. Installed the latest version of exipick from John Jetmore.
686 29. In an address list, if the pattern was not a regular expression, an empty
687 subject address (from a bounce message) matched only if the pattern was an
688 empty string. Non-empty patterns were not even tested. This was the wrong
689 because it is perfectly reasonable to use an empty address as part of a
690 database query. An empty address is now tested by patterns that are
691 lookups. However, all the other forms of pattern expect the subject to
692 contain a local part and a domain, and therefore, for them, an empty
693 address still always fails if the pattern is not itself empty.
695 30. Exim went into a mad DNS loop when attempting to do a callout where the
696 host was specified on an smtp transport, and looking it up yielded more
699 31. Re-factored the code for checking spool and log partition space into a
700 function that finds that data and another that does the check. The former
701 is then used to implement four new variables: $spool_space, $log_space,
702 $spool_inodes, and $log_inodes.
704 32. The RFC2047 encoding function was originally intended for short strings
705 such as real names; it was not keeping to the 75-character limit for
706 encoded words that the RFC imposes. It now respects the limit, and
707 generates multiple encoded words if necessary. To be on the safe side, I
708 have increased the buffer size for the ${rfc2047: expansion operator from
711 33. It is now permitted to omit both strings after an "if" condition; if the
712 condition is true, the result is "true". As before, when the second string
713 is omitted, a false condition yields an empty string. This makes it less
714 cumbersome to write custom ACL and router conditions.
716 34. Failure to deliver a bounce message always caused it to be frozen, even if
717 there was an errors_to setting on the router. The errors_to setting is now
720 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
721 canonical form (fully expanded) before being placed in
722 $sender_host_address.
724 36. The table in the code that translates DNS record types into text (T_A to
725 "A" for instance) was missing entries for NS and CNAME. It is just possible
726 that this could have caused confusion if both these types were looked up
727 for the same domain, because the text type is used as part of Exim's
728 per-process caching. But the chance of anyone hitting this buglet seems
731 37. The dnsdb lookup has been extended in a number of ways.
733 (1) There is a new type, "zns", which walks up the domain tree until it
734 finds some nameserver records. It should be used with care.
736 (2) There is a new type, "mxh", which is like "mx" except that it returns
737 just the host names, not the priorities.
739 (3) It is now possible to give a list of domains (or IP addresses) to be
740 looked up. The behaviour when one of the lookups defers can be
741 controlled by a keyword.
743 (4) It is now possible to specify the separator character for use when
744 multiple records are returned.
746 38. The dnslists ACL condition has been extended: it is now possible to supply
747 a list of IP addresses and/or domains to be looked up in a particular DNS
750 39. Added log_selector=+queue_time_overall.
752 40. When running the queue in the test harness, wait just a tad after forking a
753 delivery process, to get repeatability of debugging output.
755 41. Include certificate and key file names in error message when GnuTLS fails
756 to set them up, because the GnuTLS error message doesn't include the name
757 of the failing file when there is a problem reading it.
759 42. Allow both -bf and -bF in the same test run.
761 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
763 44. The "Exiscan patch" is now merged into the mainline Exim source.
765 45. Sometimes the final signoff response after QUIT could fail to get
766 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
767 < 0 before doing a fflush(). This bug looks as though it goes back to the
768 introduction of TLS in release 3.20, but "sometimes" must have been rare
769 because the tests only now provoked it.
771 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
772 (this can affect the format of dates).
774 47. exim_tidydb, when checking for the continued existence of a message for
775 which it has found a message-specific retry record, was not finding
776 messages that were in split spool directories. Consequently, it was
777 deleting retry records that should have stayed in existence.
779 48. Steve fixed some bugs in eximstats.
781 49. The SPA authentication driver was not abandoning authentication and moving
782 on to the next authenticator when an expansion was forced to fail,
783 contradicting the general specification for all authenticators. Instead it
784 was generating a temporary error. It now behaves as specified.
786 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
787 (the order specifies the preference for clients). The order is now AES256,
788 AES128, 3DES, ARCFOUR128.
790 51. Small patch to Sieve code - explicitly set From: when generating an
793 52. Exim crashed if a remote delivery caused a very long error message to be
794 recorded - for instance if somebody sent an entire SpamAssassin report back
795 as a large number of 550 error lines. This bug was coincidentally fixed by
796 increasing the size of one of Exim's internal buffers (big_buffer) that
797 happened as part of the Exiscan merge. However, to be on the safe side, I
798 have made the code more robust (and fixed the comments that describe what
801 53. Now that there can be additional text after "Completed" in log lines (if
802 the queue_time_overall log selector is set), a one-byte patch to exigrep
803 was needed to allow it to recognize "Completed" as not the last thing in
806 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
807 patch that reportedly fixes this has been added. I am not expert enough to
808 create a test for it. This is what the patch creator wrote:
810 "I found a little strange behaviour of ldap code when working with
811 Windows 2003 AD Domain, where users was placed in more than one
812 Organization Units. When I tried to give exim partial DN, the exit code
813 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
814 But simultaneously result of request was absolutely normal ldap result,
815 so I produce this patch..."
817 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
818 REFERENCE, so I have modified the code to exclude the patch when that macro
821 55. Some experimental protocols are using DNS PTR records for new purposes. The
822 keys for these records are domain names, not reversed IP addresses. The
823 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
824 leaves it alone. Component reversal etc. now happens only for IP addresses.
827 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
829 57. Double the size of the debug message buffer (to 2048) so that more of very
830 long debug lines gets shown.
832 58. The exicyclog utility now does better if the number of log files to keep
833 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
835 59. Two changes related to the smtp_active_hostname option:
837 (1) $smtp_active_hostname is now available as a variable.
838 (2) The default for smtp_banner uses $smtp_active_hostname instead
839 of $primary_hostname.
841 60. The host_aton() function is supposed to be passed a string that is known
842 to be a valid IP address. However, in the case of IPv6 addresses, it was
843 not checking this. This is a hostage to fortune. Exim now panics and dies
844 if the condition is not met. A case was found where this could be provoked
845 from a dnsdb PTR lookup with an IPv6 address that had more than 8
846 components; fortuitously, this particular loophole had already been fixed
847 by change 4.50/55 above.
849 If there are any other similar loopholes, the new check in host_aton()
850 itself should stop them being exploited. The report I received stated that
851 data on the command line could provoke the exploit when Exim was running as
852 exim, but did not say which command line option was involved. All I could
853 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
857 61. There was a buffer overflow vulnerability in the SPA authentication code
858 (which came originally from the Samba project). I have added a test to the
859 spa_base64_to_bits() function which I hope fixes it.
862 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
863 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
865 63. The daemon start-up calls getloadavg() while still root for those OS that
866 need the first call to be done as root, but it missed one case: when
867 deliver_queue_load_max is set with deliver_drop_privilege. This is
868 necessary for the benefit of the queue runner, because there is no re-exec
869 when deliver_drop_privilege is set.
871 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
874 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
875 was in use, was not putting the data itself into the right store pool;
876 consequently, it could be overwritten for a subsequent message in the same
877 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
880 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
882 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
883 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
884 the function were treating the return as a boolean value, which happened to
885 work because 0=false and not-0=true, but is not correct code.
887 68. The host_aton() function was not handling scoped IPv6 addresses (those
888 with, for example, "%eth0" on the end) correctly.
890 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
891 negated items (that is, ~something) in unsigned ints. Some compilers
892 apparently mutter when there is no cast.
894 70. If an address verification called from an ACL failed, and did not produce a
895 user-specific message (i.e. there was only a "system" message), nothing was
896 put in $acl_verify_message. In this situation, it now puts the system
899 71. Change 4.23/11 added synchronization checking at the start of an SMTP
900 session; change 4.31/43 added the unwanted input to the log line - except
901 that it did not do this in the start of session case. It now does.
903 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
904 This is wrong and can cause the other end to generate a synchronization
905 error if it is another Exim or anything else that does the synchronization
906 check. A QUIT command is no longer sent after a timeout.
908 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
911 74. The defer_ok option of callout verification was not working if it was used
912 when verifying addresses in header lines, that is, for this case:
914 verify = header_sender/callout=defer_ok
916 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
917 those file descriptors could be used for SMTP connections. If anything
918 wrote to stderr (the example that came up was "warn" in embedded Perl), it
919 could be sent to the SMTP client, causing chaos. The daemon now opens
920 stdin, stdout, and stderr to /dev/null when it puts itself into the
923 76. Arrange for output from Perl's "warn" command to be written to Exim's main
924 log by default. The user can override this with suitable Perl magic.
926 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
927 the log message when discard triggers, was not working for the DATA ACL or
928 for the non-SMTP ACL.
930 78. Error message wording change in sieve.c.
932 79. If smtp_accept_max_per_host was set, the number of connections could be
933 restricted to fewer than expected, because the daemon was trying to set up
934 a new connection before checking whether the processes handling previous
935 connections had finished. The check for completed processes is now done
936 earlier. On busy systems, this bug wouldn't be noticed because something
937 else would have woken the daemon, and it would have reaped the completed
940 80. If a message was submitted locally by a user whose login name contained one
941 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
942 It caused a spool format error. I have fixed the spool reading code. A
943 related problem was that the "from" clause in the Received: line became
944 illegal because of the space(s). It is now covered by ${quote_local_part.
946 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
949 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
951 83. Patch from Sieve maintainer to fix unterminated string problem in
954 84. Some minor changes to the Linux configuration files to help with other
955 OS variants using glibc.
957 85. One more patch for Sieve to update vacation handling to latest spec.
960 ----------------------------------------------------
961 See the note above about the 4.44 and 4.50 releases.
962 ----------------------------------------------------
968 1. Change 4.43/35 introduced a bug that caused file counts to be
969 incorrectly computed when quota_filecount was set in an appendfile
972 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
973 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
976 3. Give more explanation in the error message when the command for a transport
977 filter fails to execute.
979 4. There are several places where Exim runs a non-Exim command in a
980 subprocess. The SIGUSR1 signal should be disabled for these processes. This
981 was being done only for the command run by the queryprogram router. It is
982 now done for all such subprocesses. The other cases are: ${run, transport
983 filters, and the commands run by the lmtp and pipe transports.
985 5. Some older OS have a limit of 256 on the maximum number of file
986 descriptors. Exim was using setrlimit() to set 1000 as a large value
987 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
988 systems. I've change it so that if it can't get 1000, it tries for 256.
990 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
991 was an oversight, and furthermore, ever since the addition of extra
992 controls (e.g. 4.43/32), the checks on when to allow different forms of
993 "control" were broken. There should now be diagnostics for all cases when a
994 control that does not make sense is encountered.
996 7. $recipients is now available in the predata ACL (oversight).
998 8. Tidy the search cache before the fork to do a delivery from a message
999 received from the command line. Otherwise the child will trigger a lookup
1000 failure and thereby defer the delivery if it tries to use (for example) a
1001 cached ldap connection that the parent has called unbind on.
1003 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1004 of $address_data from the recipient verification was clobbered by the
1005 sender verification.
1007 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1008 was its contents. (It was OK if the option was not defined at all.)
1010 11. A "Completed" log line is now written for messages that are removed from
1011 the spool by the -Mrm option.
1013 12. $host_address is now set to the target address during the checking of
1014 ignore_target_hosts.
1016 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1017 being passed; this would have caused $sender_host_name to have been used if
1018 matching the list had actually called for a host name (not very likely,
1019 since this list is usually IP addresses). A host name is now passed as
1022 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1023 code that specifies a non-restarting handler (typically sigaction() in
1024 modern systems) in an attempt to fix a rare and obscure crash bug.
1026 15. Narrowed the window for a race in the daemon that could cause it to ignore
1027 SIGCHLD signals. This is not a major problem, because they are used only to
1028 wake it up if nothing else does.
1030 16. A malformed maildirsize file could cause Exim to calculate negative values
1031 for the mailbox size or file count. Odd effects could occur as a result.
1032 The maildirsize information is now recalculated if the size or filecount
1035 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1036 support for a long time. Removed HAVE_SYS_VFS_H.
1038 18. Updated exipick to current release from John Jetmore.
1040 19. Allow an empty sender to be matched against a lookup in an address list.
1041 Previously the only cases considered were a regular expression, or an
1044 20. Exim went into a mad DNS lookup loop when doing a callout where the
1045 host was specified on the transport, if the DNS lookup yielded more than
1048 21. The RFC2047 encoding function was originally intended for short strings
1049 such as real names; it was not keeping to the 75-character limit for
1050 encoded words that the RFC imposes. It now respects the limit, and
1051 generates multiple encoded words if necessary. To be on the safe side, I
1052 have increased the buffer size for the ${rfc2047: expansion operator from
1055 22. Failure to deliver a bounce message always caused it to be frozen, even if
1056 there was an errors_to setting on the router. The errors_to setting is now
1059 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1060 canonical form (fully expanded) before being placed in
1061 $sender_host_address.
1063 24. Updated eximstats to version 1.33
1065 25. Include certificate and key file names in error message when GnuTLS fails
1066 to set them up, because the GnuTLS error message doesn't include the name
1067 of the failing file when there is a problem reading it.
1069 26. Expand error message when OpenSSL has problems setting up cert/key files.
1072 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1073 (this can affect the format of dates).
1075 28. exim_tidydb, when checking for the continued existence of a message for
1076 which it has found a message-specific retry record, was not finding
1077 messages that were in split spool directories. Consequently, it was
1078 deleting retry records that should have stayed in existence.
1080 29. eximstats updated to version 1.35
1081 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1082 1.35 - bugfix such that pie charts by volume are generated correctly
1084 30. The SPA authentication driver was not abandoning authentication and moving
1085 on to the next authenticator when an expansion was forced to fail,
1086 contradicting the general specification for all authenticators. Instead it
1087 was generating a temporary error. It now behaves as specified.
1089 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1090 (the order specifies the preference for clients). The order is now AES256,
1091 AES128, 3DES, ARCFOUR128.
1093 31. Small patch to Sieve code - explicitly set From: when generating an
1096 32. Exim crashed if a remote delivery caused a very long error message to be
1097 recorded - for instance if somebody sent an entire SpamAssassin report back
1098 as a large number of 550 error lines. This bug was coincidentally fixed by
1099 increasing the size of one of Exim's internal buffers (big_buffer) that
1100 happened as part of the Exiscan merge. However, to be on the safe side, I
1101 have made the code more robust (and fixed the comments that describe what
1104 33. Some experimental protocols are using DNS PTR records for new purposes. The
1105 keys for these records are domain names, not reversed IP addresses. The
1106 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1107 leaves it alone. Component reversal etc. now happens only for IP addresses.
1110 34. The host_aton() function is supposed to be passed a string that is known
1111 to be a valid IP address. However, in the case of IPv6 addresses, it was
1112 not checking this. This is a hostage to fortune. Exim now panics and dies
1113 if the condition is not met. A case was found where this could be provoked
1114 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1115 components; fortuitously, this particular loophole had already been fixed
1116 by change 4.50/55 or 4.44/33 above.
1118 If there are any other similar loopholes, the new check in host_aton()
1119 itself should stop them being exploited. The report I received stated that
1120 data on the command line could provoke the exploit when Exim was running as
1121 exim, but did not say which command line option was involved. All I could
1122 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1123 running as the user.
1126 35. There was a buffer overflow vulnerability in the SPA authentication code
1127 (which came originally from the Samba project). I have added a test to the
1128 spa_base64_to_bits() function which I hope fixes it.
1131 36. The daemon start-up calls getloadavg() while still root for those OS that
1132 need the first call to be done as root, but it missed one case: when
1133 deliver_queue_load_max is set with deliver_drop_privilege. This is
1134 necessary for the benefit of the queue runner, because there is no re-exec
1135 when deliver_drop_privilege is set.
1137 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1138 was in use, was not putting the data itself into the right store pool;
1139 consequently, it could be overwritten for a subsequent message in the same
1140 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1143 38. Sometimes the final signoff response after QUIT could fail to get
1144 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1145 < 0 before doing a fflush(). This bug looks as though it goes back to the
1146 introduction of TLS in release 3.20, but "sometimes" must have been rare
1147 because the tests only now provoked it.
1153 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1154 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1155 Now it returns an integer. A number of calls were still expecting a T/F
1156 return. Fortuitously, in all cases, the tests worked in OK situations,
1157 which is the norm. However, things would have gone wrong on any write
1158 failures on the smtp file descriptor. This function is used when sending
1159 messages over SMTP and also when doing verify callouts.
1161 2. When Exim is called to do synchronous delivery of a locally submitted
1162 message (the -odf or -odi options), it no longer closes stderr before doing
1165 3. Implemented the mua_wrapper option.
1167 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1169 5. Implemented the functions header_remove(), header_testname(),
1170 header_add_at_position(), and receive_remove_recipient(), and exported them
1173 6. If an ACL "warn" statement specified the addition of headers, Exim already
1174 inserted X-ACL-Warn: at the start if there was no header name. However, it
1175 was not making this test for the second and subsequent header lines if
1176 there were newlines in the string. This meant that an invalid header could
1177 be inserted if Exim was badly configured.
1179 7. Allow an ACL "warn" statement to add header lines at the start or after all
1180 the Received: headers, as well as at the end.
1182 8. Added the rcpt_4xx retry error code.
1184 9. Added postmaster_mailfrom=xxx to callout verification option.
1186 10. Added mailfrom=xxxx to the callout verification option, for verify=
1189 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1190 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1192 12. Inserted some casts to stop certain compilers warning when using pointer
1193 differences as field lengths or precisions in printf-type calls (mostly
1194 affecting debugging statements).
1196 13. Added optional readline() support for -be (dynamically loaded).
1198 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1199 same clock tick as a message's arrival, so that its received time was the
1200 same as the "first fail" time on the retry record, and that message
1201 remained on the queue past the ultimate address timeout, every queue runner
1202 would try a delivery (because it was past the ultimate address timeout) but
1203 after another failure, the ultimate address timeout, which should have then
1204 bounced the address, did not kick in. This was a "< instead of <=" error;
1205 in most cases the first failure would have been in the next clock tick
1206 after the received time, and all would be well.
1208 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1209 being recognized when the domain list was tested by the match_domain
1210 condition in an expansion string.
1212 16. Added the ${str2b64: operator.
1214 17. Exim was always calling setrlimit() to set a large limit for the number of
1215 processes, without checking whether the existing limit was already
1216 adequate. (It did check for the limit on file descriptors.) Furthermore,
1217 errors from getrlimit() and setrlimit() were being ignored. Now they are
1218 logged to the main and panic logs, but Exim does carry on, to try to do its
1219 job under whatever limits there are.
1221 18. Imported PCRE 5.0.
1223 19. Trivial typo in log message " temporarily refused connection" (the leading
1226 20. If the log selector return_path_on_delivery was set and an address was
1227 redirected to /dev/null, the delivery process crashed because it assumed
1228 that a return path would always be set for a "successful" delivery. In this
1229 case, the whole delivery is bypassed as an optimization, and therefore no
1232 21. Internal re-arrangement: the function for sending a challenge and reading
1233 a response while authentication was assuming a zero-terminated challenge
1234 string. It's now changed to take a pointer and a length, to allow for
1235 binary data in such strings.
1237 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1239 23. Exim was not respecting finduser_retries when seeking the login of the
1240 uid under which it was called; it was always trying 10 times. (The default
1241 setting of finduser_retries is zero.) Also, it was sleeping after the final
1242 failure, which is pointless.
1244 24. Implemented tls_on_connect_ports.
1246 25. Implemented acl_smtp_predata.
1248 26. If the domain in control=submission is set empty, Exim assumes that the
1249 authenticated id is a complete email address when it generates From: or
1250 Sender: header lines.
1252 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1253 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1254 chown and chgrp in /bin and hostname in /usr/bin.
1256 28. Exim was keeping the "process log" file open after each use, just as it
1257 does for the main log. This opens the possibility of it remaining open for
1258 long periods when the USR1 signal hits a daemon. Occasional processlog
1259 errors were reported, that could have been caused by this. Anyway, it seems
1260 much more sensible not to leave this file open at all, so that is what now
1263 29. The long-running daemon process does not normally write to the log once it
1264 has entered its main loop, and it closes the log before doing so. This is
1265 so that log files can straightforwardly be renamed and moved. However,
1266 there are a couple of unusual error situations where the daemon does write
1267 log entries, and I had neglected to close the log afterwards.
1269 30. The text of an SMTP error response that was received during a remote
1270 delivery was being truncated at 512 bytes. This is too short for some of
1271 the long messages that one sometimes sees. I've increased the limit to
1274 31. It is now possible to make retry rules that apply only when a message has a
1275 specific sender, in particular, an empty sender.
1277 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1278 it possible to be selective about when SMTP synchronization is enforced.
1280 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1282 32. Implemented hosts_connection_nolog.
1284 33. Added an ACL for QUIT.
1286 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1289 35. Added mailbox_size and mailbox_filecount to appendfile.
1291 36. Added control = no_multiline_responses to ACLs.
1293 37. There was a bug in the logic of the code that waits for the clock to tick
1294 in the case where the clock went backwards by a substantial amount such
1295 that the microsecond fraction of "now" was more than the microsecond
1296 fraction of "then" (but the whole seconds number was less).
1298 38. Added support for the libradius Radius client library this is found on
1299 FreeBSD (previously only the radiusclient library was supported).
1305 1. When certain lookups returned multiple values in the form name=value, the
1306 quoting of the values was not always being done properly. Specifically:
1307 (a) If the value started with a double quote, but contained no whitespace,
1309 (b) If the value contained whitespace other than a space character (i.e.
1310 tabs or newlines or carriage returns) it was not quoted.
1311 This fix has been applied to the mysql and pgsql lookups by writing a
1312 separate quoting function and calling it from the lookup code. The fix
1313 should probably also be applied to nisplus, ibase and oracle lookups, but
1314 since I cannot test any of those, I have not disturbed their existing code.
1316 2. A hit in the callout cache for a specific address caused a log line with no
1317 reason for rejecting RCPT. Now it says "Previous (cached) callout
1318 verification failure".
1320 3. There was an off-by-one bug in the queryprogram router. An over-long
1321 return line was truncated at 256 instead of 255 characters, thereby
1322 overflowing its buffer with the terminating zero. As well as fixing this, I
1323 have increased the buffer size to 1024 (and made a note to document this).
1325 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1326 when Exim is waiting for an SMTP response from a remote server, Exim
1327 restarts its select() call on the socket, thereby resetting its timeout.
1328 This is not a problem when such interrupts are rare. Somebody set up a cron
1329 job to run exiwhat every 2 minutes, which is less than the normal select()
1330 timeout (5 or 10 minutes). This meant that the select() timeout never
1331 kicked in because it was always reset. I have fixed this by comparing the
1332 time when an interrupt arrives with the time at the start of the first call
1333 to select(). If more time than the timeout has elapsed, the interrupt is
1334 treated as a timeout.
1336 5. Some internal re-factoring in preparation for the addition of Sieve
1337 extensions (by MH). In particular, the "personal" test is moved to a
1338 separate function, and given an option for scanning Cc: and Bcc: (which is
1339 not set for Exim filters).
1341 6. When Exim created an email address using the login of the caller as the
1342 local part (e.g. when creating a From: or Sender: header line), it was not
1343 quoting the local part when it contained special characters such as @.
1345 7. Installed new OpenBSD configuration files.
1347 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1348 try to make them clearer.
1350 9. Callout options, other than the timeout value, were being ignored when
1351 verifying sender addresses in header lines. For example, when using
1353 verify = header_sender/callout=no_cache
1355 the cache was (incorrectly) being used.
1357 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1358 only the Cygwin environment.
1360 11. When return_path_on_delivery was set as a log selector, if different remote
1361 addresses in the same message used different return paths and parallel
1362 remote delivery occurred, the wrong values would sometimes be logged.
1363 (Whenever a remote delivery process finished, the return path value from
1364 the most recently started remote delivery process was logged.)
1366 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1367 lines when AUTH and/or TLS are in use. This is the "received protocol"
1368 field. Exim used to use "asmtp" for authenticated SMTP, without any
1369 indication (in the protocol name) for TLS use. Now it follows the RFC and
1370 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1371 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1372 names appear in log lines as well as in Received: header lines.
1374 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1375 extensions, and comparison tests, and to fix some bugs.
1377 14. Changes to the "personal" filter test:
1379 (1) The test was buggy in that it was just doing the equivalent of
1380 "contains" tests on header lines. For example, if a user's address was
1381 anne@some.where, the "personal" test would incorrectly be true for
1383 To: susanne@some.where
1385 This test is now done by extracting each address from the header in turn,
1386 and checking the entire address. Other tests that are part of "personal"
1387 are now done using regular expressions (for example, to check local parts
1388 of addresses in From: header lines).
1390 (2) The list of non-personal local parts in From: addresses has been
1391 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1392 taken from the Sieve specification recommendations.
1394 (3) If the message contains any header line starting with "List-" it is
1395 treated as non-personal.
1397 (4) The test for "circular" in the Subject: header line has been removed
1398 because it now seems ill-conceived.
1400 15. Minor typos in src/EDITME comments corrected.
1402 16. Installed latest exipick from John Jetmore.
1404 17. If headers_add on a router specified a text string that was too long for
1405 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1406 of string_sprintf() is now avoided.
1408 18. $message_body_size was not set (it was always zero) when running the DATA
1409 ACL and the local_scan() function.
1411 19. For the "mail" command in an Exim filter, no default was being set for
1412 the once_repeat time, causing a random time value to be used if "once" was
1413 specified. (If the value happened to be <= 0, no repeat happened.) The
1414 default is now 0s, meaning "never repeat". The "vacation" command was OK
1415 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1416 (I found it when inspecting the code).
1418 20. There is now an overall timeout for performing a callout verification. It
1419 defaults to 4 times the callout timeout, which applies to individual SMTP
1420 commands during the callout. The overall timeout applies when there is more
1421 than one host that can be tried. The timeout is checked before trying the
1422 next host. This prevents very long delays if there are a large number of
1423 hosts and all are timing out (e.g. when the network connections are timing
1424 out). The value of the overall timeout can be changed by specifying an
1425 additional sub-option for "callout", called "maxwait". For example:
1427 verify = sender/callout=5s,maxwait=20s
1429 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1430 to the end before writing, but this should make it even safer).
1432 22. Exim was forgetting that it had advertised PIPELINING for the second and
1433 subsequent messages on an SMTP connection. It was also not resetting its
1434 memory on STARTTLS and an internal HELO.
1436 23. When Exim logs an SMTP synchronization error within a session, it now
1437 records whether PIPELINING has been advertised or not.
1439 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1440 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1441 rather than long int.
1443 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1445 26. Added the never_mail option to autoreply.
1451 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1452 crash if the getsockname() call failed; this can happen if a connection is
1453 closed very soon after it is established. The problem was simply in the
1454 order in which certain operations were done, causing Exim to try to write
1455 to the SMTP stream before it had set up the file descriptor. The bug has
1456 been fixed by making things happen in the correct order.
1462 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1463 before the connection was closed, thus losing the rejection response.
1465 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1466 some early Solaris releases, but causes trouble in current releases where
1467 socklen_t is defined.
1469 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1472 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1475 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1476 that releasing the top of it at the end releases what was used for sub-
1477 expansions (unless the block got too big). However, discard this block if
1478 the first thing is a variable or header, so that we can use its block when
1479 it is dynamic (useful for very large $message_headers, for example).
1481 6. Lookups now cache *every* query, not just the most recent. A new, separate
1482 store pool is used for this. It can be recovered when all lookup caches are
1483 flushed. Lookups now release memory at the end of their result strings.
1484 This has involved some general refactoring of the lookup sources.
1486 7. Some code has been added to the store_xxx() functions to reduce the amount
1487 of flapping under certain conditions.
1489 8. log_incoming_interface used to affect only the <= reception log lines. Now
1490 it causes the local interface and port to be added to several more SMTP log
1491 lines, for example "SMTP connection from", and rejection lines.
1493 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1495 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1497 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1498 could be overwritten at the end of the current message (or the start of a
1499 new message if it was set in a HELO ACL). The value is now preserved for
1500 the duration of the SMTP connection.
1502 12. If a transport had a headers_rewrite setting, and a matching header line
1503 contained an unqualified address, that address was qualified, even if it
1504 did not match any rewriting rules. The underlying bug was that the values
1505 of the flags that permit the existence of unqualified sender and recipient
1506 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1507 non-local messages, and by -bnq for local messages) were not being
1508 preserved with the message after it was received.
1510 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1511 "next input=" as part of the text comprising the host identity instead of
1512 the correct text. The code was using the same buffer for two different
1513 strings. However, depending on which order the printing function evaluated
1514 its arguments, the bug did not always show up. Under Linux, for example, my
1515 test suite worked just fine.
1517 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1518 doesn't work with some older versions of Perl. It has been changed to "my",
1519 which in any case is probably the better facility to use.
1521 15. A really picky compiler found some instances of statements for creating
1522 error messages that either had too many or two few arguments for the format
1525 16. The size of the buffer for calls to the DNS resolver has been increased
1526 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1527 for addresses that have a lot of PTR records. This alleviates a problem; it
1528 does not fully solve it.
1530 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1531 buffer now truncates the list and logs the incident, which is the same
1532 action as happens when Exim is looking up a host name and its aliases.
1533 Previously in this situation something unpredictable would happen;
1534 sometimes it was "internal error: store_reset failed".
1536 18. If a server dropped the connection unexpectedly when an Exim client was
1537 using GnuTLS and trying to read a response, the client delivery process
1538 crashed while trying to generate an error log message.
1540 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1541 string, for example:
1543 warn message = H1: something\nH2: something
1545 the text was added as a single header line from Exim's point of view
1546 though it ended up OK in the delivered message. However, searching for the
1547 second and subsequent header lines using $h_h2: did not work. This has been
1548 fixed. Similarly, if a system filter added multiple headers in this way,
1549 the routers could not see them.
1551 20. Expanded the error message when iplsearch is called with an invalid key to
1552 suggest using net-iplsearch in a host list.
1554 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1555 ACLs are no longer actually imposed (and a message to that effect is
1558 22. If a "gecos" field in a passwd entry contained escaped characters, in
1559 particular, if it contained a \" sequence, Exim got it wrong when building
1560 a From: or a Sender: header from that name. A second bug also caused
1561 incorrect handling when an unquoted " was present following a character
1562 that needed quoting.
1564 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1565 was not being matched caselessly.
1567 24. Arranged for all hyphens in the exim.8 source to be escaped with
1570 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1571 properly. Recipient callout cache records were still being keyed to include
1572 the sender, even when use_sender was set false. This led to far more
1573 callouts that were necessary. The sender is no longer included in the key
1574 when use_sender is false.
1576 26. Added "control = submission" modifier to ACLs.
1578 27. Added the ${base62d: operator to decode base 62 numbers.
1580 28. dnsdb lookups can now access SRV records.
1582 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1583 the configuration file.
1585 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1586 (-v) mode. This makes the output for a verbose queue run more intelligible.
1588 31. Added a use_postmaster feature to recipient callouts.
1590 32. Added the $body_zerocount variable, containing the number of binary zero
1591 bytes in the message body.
1593 33. The time of last modification of the "new" subdirectory is now used as the
1594 "mailbox time last read" when there is a quota error for a maildir
1597 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1599 35. Added +ignore_unknown as a special item in host lists.
1601 36. Code for decoding IPv6 addresses in host lists is now included, even if
1602 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1603 address was recognized as an IP address, but was then not correctly decoded
1604 into binary, causing unexpected and incorrect effects when compared with
1611 1. Very minor rewording of debugging text in manualroute to say "list of
1612 hosts" instead of "hostlist".
1614 2. If verify=header_syntax was set, and a header line with an unqualified
1615 address (no domain) and a large number of spaces between the end of the
1616 name and the colon was received, the reception process suffered a buffer
1617 overflow, and (when I tested it) crashed. This was caused by some obsolete
1618 code that should have been removed. The fix is to remove it!
1620 3. When running in the test harness, delay a bit after writing a bounce
1621 message to get a bit more predictability in the log output.
1623 4. Added a call to search_tidyup() just before forking a reception process. In
1624 theory, someone could use a lookup in the expansion of smtp_accept_max_
1625 per_host which, without the tidyup, could leave open a database connection.
1627 5. Added the variables $recipient_data and $sender_data which get set from a
1628 lookup success in an ACL "recipients" or "senders" condition, or a router
1629 "senders" option, similar to $domain_data and $local_part_data.
1631 6. Moved the writing of debug_print from before to after the "senders" test
1634 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1635 problems for message scanning, either using a data ACL, or using
1636 local_scan() because the Received: header was not generated till after they
1637 were called (in order to set the time as the time of reception completion).
1638 I have revised the way this works. The header is now generated after the
1639 body is received, but before the ACL or local_scan() are called. After they
1640 are run, the timestamp in the header is updated.
1646 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1647 before starting a queue runner without re-exec. This happened only when
1648 deliver_drop_privilege was set or when the Exim user was set to root. The
1649 effect of the bug was that timeouts during subsequent deliveries caused
1650 crashes instead of being properly handled. The handler is now left at its
1651 default (and expected) setting.
1653 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1654 message, again when deliver_drop_privilege is set or Exim is run as root.
1655 The bug described in (1) was not present in this case, but the tidying up
1656 of the other signals was missing. I have made the two cases consistent.
1658 3. The ignore_target_hosts setting on a manualroute router was being ignored
1659 for hosts that were looked up using the /MX notation.
1661 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1664 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1665 operated on the sender address. After changing the $sender_address to <>
1666 for the sender address verify, Exim was re-instated it as the original
1667 (before rewriting) address, but remembering that it had rewritten it, so it
1668 wasn't rewriting it again. This bug also had the effect of breaking the
1669 sender address verification caching when the sender address was rewritten.
1671 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1672 This has been changed so that if the ip literal address matches
1673 ignore_target_hosts, the router declines.
1675 7. Added expansion conditions match_domain, match_address, and match_local_
1676 part (NOT match_host).
1678 8. The placeholder for the Received: header didn't have a length field set.
1680 9. Added code to Exim itself and to exim_lock to test for a specific race
1681 condition that could lead to file corruption when using MBX delivery. The
1682 issue is with the lockfile that is created in /tmp. If this file is removed
1683 after a process has opened it but before that process has acquired a lock,
1684 there is the potential for a second process to recreate the file and also
1685 acquire a lock. This could lead to two Exim processes writing to the file
1686 at the same time. The added code performs the same test as UW imapd; it
1687 checks after acquiring the lock that its file descriptor still refers to
1688 the same named file.
1690 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1691 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1694 11. Added the smtp_active_hostname option. If used, this will typically be made
1695 to depend on the incoming interface address. Because $interface_address is
1696 not set up until the daemon has forked a reception process, error responses
1697 that can happen earlier (such as "too many connections") no longer contain
1700 12. If an expansion in a condition on a "warn" statement fails because a lookup
1701 defers, the "warn" statement is abandoned, and the next ACL statement is
1702 processed. Previously this caused the whole ACL to be aborted.
1704 13. Added the iplsearch lookup type.
1706 14. Added ident_timeout as a log selector.
1708 15. Added tls_certificate_verified as a log selector.
1710 16. Added a global option tls_require_ciphers (compare the smtp transport
1711 option of the same name). This controls incoming TLS connections.
1713 17. I finally figured out how to make tls_require_ciphers do a similar thing
1714 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1715 before starting the TLS session.
1717 18. Tabs are now shown as \t in -bP output.
1719 19. If the log selector return_path_on_delivery was set, Exim crashed when
1720 bouncing a message because it had too many Received: header lines.
1722 20. If two routers both had headers_remove settings, and the first one included
1723 a superfluous trailing colon, the final name in the first list and the
1724 first name in the second list were incorrectly joined into one item (with a
1725 colon in the middle).
1731 1. Added -C and -D options to the exinext utility, mainly to make it easier
1732 to include in the automated testing, but these could be helpful when
1733 multiple configurations are in use.
1735 2. The exinext utility was not formatting the output nicely when there was
1736 an alternate port involved in the retry record key, nor when there was a
1737 message id as well (for retries that were specific to a specific message
1738 and a specific host). It was also confused by IPv6 addresses, because of
1739 the additional colons they contain. I have fixed the IPv4 problem, and
1740 patched it up to do a reasonable job for IPv6.
1742 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1743 delivery, the log line now contains "pipelined" if PIPELINING was used.
1745 4. An SMTP transport process used to panic and die if the bind() call to set
1746 an explicit outgoing interface failed. This has been changed; it is now
1747 treated in the same way as a connect() failure.
1749 5. A reference to $sender_host_name in the part of a conditional expansion
1750 that was being skipped was still causing a DNS lookup. This no longer
1753 6. The def: expansion condition was not recognizing references to header lines
1754 that used bh_ and bheader_.
1756 7. Added the _cache feature to named lists.
1758 8. The code for checking quota_filecount in the appendfile transport was
1759 allowing one more file than it should have been.
1761 9. For compatibility with Sendmail, the command line option
1769 and sets the incoming protocol and host name (for trusted callers). The
1770 host name and its colon can be omitted when only the protocol is to be set.
1771 Note the Exim already has two private options, -pd and -ps, that refer to
1772 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1773 "s", but I don't think that's a major issue.
1775 10. A number of refactoring changes to the code, none of which should affect
1778 (a) The number of logging options was getting close to filling up the
1779 32-bit word that was used as a bit map. I have split them into two classes:
1780 those that are passed in the argument to log_write(), and those that are
1781 only ever tested independently outside of that function. These are now in
1782 separate 32-bit words, so there is plenty of room for expansion again.
1783 There is no change in the user interface or the logging behaviour.
1785 (b) When building, for example, log lines, the code previously used a
1786 macro that called string_cat() twice, in order to add two strings. This is
1787 not really sufficiently general. Furthermore, there was one instance where
1788 it was actually wrong because one of the argument was used twice, and in
1789 one call a function was used. (As it happened, calling the function twice
1790 did not affect the overall behaviour.) The macro has been replaced by a
1791 function that can join an arbitrary number of extra strings onto a growing
1794 (c) The code for expansion conditions now uses a table and a binary chop
1795 instead of a serial search (which was left over from when there were very
1796 few conditions). Also, it now recognizes conditions like "pam" even when
1797 the relevant support is not compiled in: a suitably worded error message is
1798 given if an attempt is made to use such a condition.
1800 11. Added ${time_interval:xxxxx}.
1802 12. A bug was causing one of the ddress fields not to be passed back correctly
1803 from remote delivery subprocesses. The field in question was not being
1804 subsequently used, so this caused to problems in practice.
1806 13. Added new log selectors queue_time and deliver_time.
1808 14. Might have fixed a bug in maildirsizefile handling that threw up
1809 "unexpected character" debug warnings, and recalculated the data
1810 unnecessarily. In any case, I expanded the warning message to give more
1813 15. Added the message "Restricted characters in address" to the statements in
1814 the default ACL that block characters like @ and % in local parts.
1816 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1817 Three changes have been made:
1819 (a) There was a serious bug; a negative response to MAIL caused the whole
1820 recipient domain to be cached as invalid, thereby blocking all messages
1821 to all local parts at the same domain, from all senders. This bug has
1822 been fixed. The domain is no longer cached after a negative response to
1823 MAIL if the sender used is not empty.
1825 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1828 (c) A new callout option, "use_sender" has been added for people who want
1829 the modified behaviour.
1835 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1838 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1839 indeed breaks things for older releases.
1841 3. Added additional logging to the case where there is a problem reading data
1842 from a filter that is running in a subprocess using a pipe, in order to
1843 try to track down a specific problem.
1845 4. Testing facility fudge: when running in the test harness and attempting
1846 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1847 getting "No route to host". Convert this to a timeout.
1849 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1852 6. Some OS don't have socklen_t but use size_t instead. This affects the
1853 fifth argument of getsockopt() amongst other things. This is now
1854 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1855 can be set for individual OS. I have set it for SunOS5, OSF1, and
1856 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1857 some earlier ones do not.
1859 7. Change 4.30/15 was not doing the test caselessly.
1861 8. The standard form for an IPv6 address literal was being rejected by address
1862 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1863 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1864 this, as well as the form without the "IPv6" on the front (but only when
1865 address literals are enabled, of course).
1867 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1869 10. Exim crashed if a message with an empty sender address specified by -f
1870 encountered a router with an errors_to setting. This could be provoked only
1871 by a command such as
1875 where an empty string was supplied; "<>" did not hit this bug.
1877 11. Installed PCRE release 4.5.
1879 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1880 remained set. It is now erased.
1882 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1883 times from message ids (which are base 36 rather than the normal 62).
1885 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1886 were being counted as actual protocol errors, and logged if the log
1887 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1888 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1889 and DATA following a set of rejected RCPTs do not count as protocol errors.
1890 In other words, Exim assumes they were pipelined, though this may not
1891 actually be the case. Of course, in all cases the client gets an
1892 appropriate error code.
1894 15. If a lookup fails in an ACL condition, a message about the failure may
1895 be available; it is used if testing the ACL cannot continue, because most
1896 such messages specify what the cause of the deferral is. However, some
1897 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1898 that caused an old message to be retained and used if a later statement
1899 caused a defer, replacing the real cause of the deferral.
1901 16. If an IP address had so many PTR records that the DNS lookup buffer
1902 was not large enough to hold them, Exim could crash while trying to process
1903 the truncated data. It now detects and logs this case.
1905 17. Further to 4.21/58, another change has been made: if (and only if) the
1906 first line of a message (the first header line) ends with CRLF, a bare LF
1907 in a subsequent header line has a space inserted after it, so as not to
1908 terminate the header.
1910 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1911 unnecessarily, used atoi() instead of strtol(), and didn't check the
1912 termination when getting file sizes from file names by regex.
1914 19. Completely re-implemented the support for maildirsize files, in the light
1915 of a number of problems with the previous contributed implementation
1916 (4.30/29). In particular:
1918 . If the quota is zero, the maildirsize file is maintained, but no quota is
1921 . If the maildir directory does not exist, it is created before any attempt
1922 to write a maildirsize file.
1924 . The quota value in the file is just a cache; if the quota is changed in
1925 the transport, the new value overrides.
1927 . A regular expression is available for excluding directories from the
1930 20. The autoreply transport checks the characters in options that define the
1931 message's headers; it allows continued headers, but it was checking with
1932 isspace() after an embedded newline instead of explicitly looking for a
1935 21. If all the "regular" hosts to which an address was routed had passed their
1936 expiry times, and had not reached their retry times, the address was
1937 bounced, even if fallback hosts were defined. Now Exim should go on to try
1940 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1941 equivalent code in the SMTP transport. Some hosts send humungous responses
1942 to HELO/EHLO, more than 1024 it seems.
1944 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1945 gives compiler warnings in some environments. I've now done it "properly",
1948 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1949 (because of problems with the built-in one) was declared to return uschar *
1950 instead of char *, causing compiler failure.
1952 25. Fixed a file descriptor leak when processing alias/forward files.
1954 26. Fixed a minor format string issue in dbfn.c.
1956 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1958 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1959 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1962 29. When a sender address is verified, it is cached, to save repeating the test
1963 when there is more than one recipient in a message. However, when the
1964 verification involves a callout, it is possible for different callout
1965 options to be set for different recipients. It is too complicated to keep
1966 track of this in the cache, so now Exim always runs a verification when a
1967 callout is required, relying on the callout cache for the optimization.
1968 The overhead is duplication of the address routing, but this should not be
1971 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
1972 to be verified with callout=postmaster, and the main callout worked but the
1973 postmaster check failed, the verification correctly failed. However, if a
1974 subsequent RCPT command asked for sender verification *without* the
1975 postmaster check, incorrect caching caused this verification also to fail,
1978 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1979 it was not caching the DNS options (qualify_single, search_parents) that
1980 were used when the lookup failed. A subsequent lookup with different
1981 options therefore always gave the same answer, though there were cases
1982 where it should not have. (Example: a "domains = !$mx_any" option on a
1983 dnslookup router: the "domains" option is always processed without any
1984 widening, but the router might have qualify_single set.) Now Exim uses the
1985 cached value only when the same options are set.
1987 32. Added John Jetmore's "exipick" utility to the distribution.
1989 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
1990 than a timeout (e.g. a certificate is required, and is not provided), an
1991 Exim server now closes the connection immediately. Previously it waited for
1992 the client to close - but if the client is SSL, it seems that they each
1993 wait for each other, leading to a delay before one of them times out.
1995 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
1996 maintained 0.8.x compatibility because I don't think many are using it, and
1997 it is clearly obsolete.
1999 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2002 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2003 client certificate was expired. A simple patch fixes this, though I don't
2004 understand the full logic of why the verify callback is called multiple
2007 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2008 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2009 which causes problems with some clients (such as the Certicom SSL Plus
2010 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2011 disables the coutermeasure allowing Eudora to connect."
2013 38. Exim was not checking that a write() to a log file succeeded. This could
2014 lead to Bad Things if a log got too big, in particular if it hit a file
2015 size limit. Exim now panics and dies if it cannot write to a log file, just
2016 as it does if it cannot open a log file.
2018 39. Modified OS/Makefile-Linux so that it now contains
2020 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2022 The two -D definitions ensure that Exim is compiled with large file
2023 support, which makes it possible to handle log files that are bigger than
2026 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2027 instance) a domain was checked against a named list that involved a lookup,
2028 causing $domain_data to be set, then another domain was checked against the
2029 same list, then the first domain was re-checked, the value of $domain_data
2030 after the final check could be wrong. In particular, if the second check
2031 failed, it could be set empty. This bug probably also applied to
2034 41. The strip_trailing_dot option was not being applied to the address given
2035 with the -f command-line option.
2037 42. The code for reading a message's header from the spool was incrementing
2038 $received_count, but never initializing it. This meant that the value was
2039 incorrect (doubled) while delivering a message in the same process in which
2040 it was received. In the most common configuration of Exim, this never
2041 happens - a fresh exec is done - but it can happen when
2042 deliver_drop_privilege is set.
2044 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2045 it now includes up to 150 characters of the unexpected data in the log
2048 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2049 and building data strings. The size of both of these buffers was 10 000
2050 bytes - far larger than anybody would *ever* want, thought I. Needless to
2051 say, somebody hit the limit. I have increased the maximum line length to
2052 20 000 and the maximum data length of concatenated lines to 100 000. I have
2053 also fixed two bugs, because there was no checking on these buffers. Tsk,
2054 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2055 buffer is too small.
2057 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2058 lsearch lookups. Now it does.
2060 46. When parsing a route_list item in a manualroute router, a fixed-length
2061 buffer was used for the list of hosts. I made this 1024 bytes long,
2062 thinking that nobody would ever have a list of hosts that long. Wrong.
2063 Somebody had a whole pile of complicated expansion conditions, and the
2064 string was silently truncated, leading to an expansion error. It turns out
2065 that it is easier to change to an unlimited length (owing to other changes
2066 that have happened since this code was originally written) than to build
2067 structure for giving a limitation error. The length of the item that
2068 expands into the list of hosts is now unlimited.
2070 47. The lsearch lookup could not handle data where the length of text line was
2071 more than 4095 characters. Such lines were truncated, leading to shortened
2072 data being returned. It should now handle lines of any length.
2074 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2075 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2078 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2079 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2080 no longer happens in comment lines. A list of replacements is now placed
2081 at the head of all of the source files, except those whose only change is
2082 to replace PERL_COMMAND in the very first #! line.
2084 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2085 messages on the queue, with a bottom-up merge sort, using code contributed
2086 by Michael Haardt. This should make operations like -bp somewhat faster on
2087 large queues. It won't affect queue runners, except when queue_run_in_order
2090 51. Installed eximstats 1.31 in the distribution.
2092 52. Added support for SRV lookups to the dnslookup router.
2094 53. If an ACL referred to $message_body or $message_body_end, the value was not
2095 reset for any messages that followed in the same SMTP session.
2097 54. The store-handling optimization for building very long strings was not
2098 differentiating between the different store pools. I don't think this
2099 actually made any difference in practice, but I've tidied it.
2101 55. While running the routers to verify a sender address, $sender_address
2102 was still set to the sender address. This is wrong, because when routing to
2103 send a bounce to the sender, it would be empty. Therefore, I have changed
2104 it so that, while verifying a sender address, $sender_address is set to <>.
2105 (There is no change to what happens when verifying a recipient address.)
2107 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2108 target A or AAAA records (if not already returned) without resetting the
2109 qualify_single or search_parents options of the DNS resolver. These are
2110 inappropriate in this case because the targets of MX and SRV records must
2111 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2112 target that, when qualified, matched something in the local domain. These
2113 two options are now turned off when doing these lookups.
2115 57. It seems that at least some releases of Reiserfs (which does not have the
2116 concept of a fixed number of inodes) returns zero and not -1 for the
2117 number of available inodes. This interacted badly with check_spool_inodes,
2118 which assumed that -1 was the "no such thing" setting. What I have done is
2119 to check that the total number of inodes is greater than zero before doing
2120 the test of how many are available.
2122 58. When a "warn" ACL statement has a log_message modifier, the message is
2123 remembered, and not repeated. This is to avoid a lot of repetition when a
2124 message has many recipients that cause the same warning to be written.
2125 Howewer, Exim was preserving the list of already written lines for an
2126 entire SMTP session, which doesn't seem right. The memory is now reset if a
2127 new message is started.
2129 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2130 debugging output unless log_rewrite was also set.
2132 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2133 of (int)(handle) when we know that handle contains (void *)(-1).
2135 61. The Exim daemon panic-logs an error return when it closes the incoming
2136 connection. However "connection reset by peer" seems to be common, and
2137 isn't really an error worthy of noting specially, so that particular error
2140 62. When Exim is trying to find all the local interfaces, it used to panic and
2141 die if the ioctl to get the interface flags failed. However, it seems that
2142 on at least one OS (Solaris 9) it is possible to have an interface that is
2143 included in the list of interfaces, but for which you get a failure error
2144 for this call. This happens when the interface is not "plumbed" into a
2145 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2146 failure of the "get flags" call assumes that the interface is down.
2148 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2149 makes life easier for people who are doing arithmetic on fields extracted
2150 from dates, where you often get leading zeros that should not be
2151 interpreted as octal.
2153 64. Added qualify_domain to the redirect router, to override the global
2156 65. If a pathologically long header line contained very many addresses (the
2157 report of this problem mentioned 10 000) and each of them was rewritten,
2158 Exim could use up a very large amount of memory. (It kept on making new
2159 copies of the header line as it rewrote, and never released the old ones.)
2160 At the expense of a bit more processing, the header rewriting function has
2161 been changed so that it no longer eats memory in this way.
2163 66. The generation of the Received: header has been moved from the time that a
2164 message starts to be received, to the time that it finishes. The timestamp
2165 in the Received: header should now be very close to that of the <= log
2166 line. There are two side-effects of this change:
2168 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2169 logged header lines no longer include the local Received: line, because
2170 it has not yet been created. The same applies to a copy of the message
2171 that is returned to a non-SMTP sender when a message is rejected.
2173 (b) When a filter file is tested using -bf, no additional Received: header
2174 is added to the test message. After some thought, I decided that this
2177 This change does not affect the value of $received_for. It is still set
2178 after address rewriting, but before local_scan() is called.
2180 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2182 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2183 gave an unhelpful panic error message, and a defer error. I have managed to
2184 change this behaviour so that it now rejects any supplied certificate,
2185 which seems right, as the list of acceptable certificates is empty.
2187 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2188 gave an unhelpful defer error. I have not managed to make this reject any
2189 supplied certificates, but the error message it gives is "no certificate
2190 supplied", which is not helpful.
2192 70. exigrep's output now also includes lines that are not associated with any
2193 message, but which match the given pattern. Implemented by a patch from
2194 Martin Sluka, which also tidied up the Perl a bit.
2196 71. Recipient callout verification, like sender verification, was using <> in
2197 the MAIL FROM command. This isn't really the right thing, since the actual
2198 sender may affect whether the remote host accepts the recipient or not. I
2199 have changed it to use the actual sender in the callout; this means that
2200 the cache record is now keyed on a recipient/sender pair, not just the
2201 recipient address. There doesn't seem to be a real danger of callout loops,
2202 since a callout by the remote host to check the sender would use <>.
2203 [SEE ABOVE: changed after hitting problems.]
2205 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2206 temporary errors. However, in the case of such a code being given after
2207 the end of a data transmission (i.e. after ".") Exim was failing to write
2208 a retry record for the message. (Yes, there was some broken host that was
2209 actually sending 8xx at this point.)
2211 73. An unknown lookup type in a host list could cause Exim to panic-die when
2212 the list was checked. (An example that provoked this was putting <; in the
2213 middle of a list instead of at the start.) If this happened during a DATA
2214 ACL check, a -D file could be left lying around. This kind of configuration
2215 error no longer causes Exim to die; instead it causes a defer errror. The
2216 incident is still logged to the main and panic logs.
2218 74. Buglet left over from Exim 3 conversion. The message "too many messages
2219 in one connection" was written to the rejectlog but not the mainlog, except
2220 when address rewriting (yes!) was being logged.
2222 75. Added write_rejectlog option.
2224 76. When a system filter was run not as root (that is, when system_filter_user
2225 was set), the values of the $n variables were not being returned to the
2226 main process; thus, they were not subsequently available in the $sn
2229 77. Added +return_path_on_delivery log selector.
2231 78. A connection timeout was being treated differently from recipients deferred
2232 when testing hosts_max_try with a message that was older than the host's
2233 retry timeout. (The host should not be counted, thus allowing all hosts to
2234 be tried at least once before bouncing.) This may have been the cause of an
2235 occasionally reported bug whereby a message would remain on the queue
2236 longer than the retry timeout, but would be bounced if a delivery was
2237 forced. I say "may" because I never totally pinned down the problem;
2238 setting up timeout/retry tests is difficult. See also the next item.
2240 79. The ultimate address timeout was not being applied to errors that involved
2241 a combination of host plus message (for example, a timeout on a MAIL
2242 command). When an address resolved to a number of possible hosts, and they
2243 were not all tried for each delivery (e.g. because of hosts_max_try), a
2244 message could remain on the queue longer than the retry timeout.
2246 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2249 81. Fixed an obscure SMTP outgoing bug which required at least the following
2250 conditions: (a) there was another message waiting for the same server;
2251 (b) the server returned 5xx to all RCPT commands in the first message so
2252 that the message was not completed; (c) the server dropped the connection
2253 or gave a negative response to the RSET that Exim sends to abort the
2254 transaction. The observed case was a dropped connection after DATA that had
2255 been sent in pipelining mode. That is, the server had advertised PIPELINING
2256 but was not implementing it correctly. The effect of the bug was incorrect
2257 behaviour, such as trying another host, and this could lead to a crash.
2263 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2264 and daemon.c were passed as pointers to ints; they should have been
2265 pointers to socklen_t variables (which are typically unsigned ints).
2267 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2270 3. Fixed a really odd bug that affected only the testing scheme; patching a
2271 certain fixed string in the binary changed the value of another string that
2272 happened to be identical to the end of the original first string.
2274 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2275 name", it returns that address as the IP address. On some operating
2276 systems (e.g. Solaris), it also passes back the IP address string as the
2277 "host name". However, on others (e.g. Linux), it passes back an empty
2278 string. Exim wasn't checking for this, and was changing the host name to an
2279 empty string, assuming it had been canonicized.
2281 5. Although rare, it is permitted to have more than one PTR record for a given
2282 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2283 all the names associated with an address, because they do in Solaris.
2284 However, it seems that they do not in Linux for data that comes from the
2285 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2286 I found this out when I moved to a new Linux workstation and tried to run
2287 the Exim test suite.
2289 To get round this problem I have changed the code so that it now does its
2290 own call to the DNS to look up PTR records when searching for a host name.
2291 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2292 addresses that are only in /etc/hosts are still found.
2294 This behaviour is, however, controlled by an option called host_lookup_
2295 order, which defaults to "bydns:byaddr". If people want to use the other
2296 order, or indeed, just use one or the other means of lookup, they can
2297 specify it in this variable.
2299 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2300 some operating systems, this comes back from gethostbyaddr() as an empty
2301 string, and this is what Exim used to test for. However, it seems that in
2302 other systems, "." is yielded. Exim now tests for this case too.
2304 7. The values of check_spool_space and check_log_space are now held internally
2305 as a number of kilobytes instead of an absolute number of bytes. If a
2306 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2307 kilobyte. This means that much larger values can be stored.
2309 8. Exim monitor: an attempt to get the action menu when not actually pointing
2310 at a message produces an empty menu entitled "No message selected". This
2311 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2312 no entries in it ("Shell widget menu has zero width and/or height"). So I
2313 have added a single, blank menu entry in this case.
2315 9. Added ${quote_local_part.
2317 10. MIME decoding is now applied to the contents of Subject: header lines when
2320 11. Now that a reference to $sender_host_address automatically causes a reverse
2321 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2322 host lookup before query-style lookups in lists that might use this
2323 variable. This has therefore been abolished, and the "net-" prefix is no
2324 longer necessary for query-style lookups.
2326 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2327 have been a typo for LFLAGS, so it has been changed.
2329 13. The install script calls Exim with "-C /dev/null" in order to find the
2330 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2331 to be output. Howeve, since Exim outputs its version number before the
2332 error, it didn't break the script. It just looked ugly. I fixed this by
2333 always allowing "-C /dev/null" if the caller is root.
2335 14. Ignore overlarge ACL variable number when reading spool file - insurance
2336 against a later release with more variables having written the file.
2338 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2339 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2340 this, as well as the form without the "IPv6" on the front.
2342 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2343 OS/Makefile-Darwin file.
2345 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2346 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2348 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2349 variable. However, some OS use ints for inodes. I've added cast to long int
2350 to get rid of the compiler warning.
2352 19. I had forgotten to lock out "/../" in configuration file names when
2353 ALT_CONFIG_PREFIX was set.
2355 20. Routers used for verification do not need to specify transports. However,
2356 if such a router generated a host list, and callout was configured, Exim
2357 crashed, because it could not find a port number from the (non-existent)
2358 transport. It now assumes port 25 in this circumstance.
2360 21. Added the -t option to exigrep.
2362 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2363 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2364 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2365 them is set, LOOKUP_LSEARCH is forced.
2367 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2369 24. Added sender and host information to the "rejected by local_scan()" log
2370 line; previously there was no indication of these.
2372 25. Added .include_if_exists.
2374 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2375 Linux. It turns out that not all file systems support this. Apparently some
2376 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2377 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2378 not supported on the file descriptor, is now ignored when Exim is trying to
2379 sync a directory. This applies only to Linux.
2381 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2383 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2384 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2387 29. Added support for maildirsize files from supplied patch (modified a bit).
2389 30. The use of :fail: followed by an empty string could lead Exim to respond to
2390 sender verification failures with (e.g.):
2392 550 Verification failed for <xxx>
2393 550 Sender verify failed
2395 where the first response line was missing the '-' that indicates it is not
2396 the final line of the response.
2398 31. The loop for finding the name of the user that called Exim had a hardwired
2399 limit of 10; it now uses the value of finduser_retries, which is used for
2400 all other user lookups.
2402 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2405 33. Exim was neglecting to zero errno before one call of strtol() when
2406 expanding a string and expecting an integer value. On some systems this
2407 resulted in spurious "integer overflow" errors. Also, it was casting the
2408 result into an int without checking.
2410 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2411 did not work. The code looks as if it has *never* worked, though it appears
2412 to have been documented since at least releast 1.62. I have made it work.
2414 35. The "timeout_DNS" error in retry rules, also documented since at least
2415 1.62, also never worked. As it isn't clear exactly what this means, and
2416 clearly it isn't a major issue, I have abolished the feature by treating it
2417 as "timeout", and writing a warning to the main and panic logs.
2419 36. The display of retry rules for -brt wasn't always showing the error code
2422 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2423 timeout_connect_A, timeout_connect_MX.
2425 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2426 to the empty sender.
2428 39. The daemon was not analysing the content of -oX till after it had closed
2429 stderr and disconnected from the controlling terminal. This meant that any
2430 syntax errors were only noted on the panic log, and the return code from
2431 the command was 0. By re-arranging the code a little, I've made the
2432 decoding happen first, so such errors now appear on stderr, and the return
2433 code is 1. However, the actual setting up of the sockets still happens in
2434 the disconnected process, so errors there are still only recorded on the
2437 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2438 connections (as happens on some IP stacks) was logged at start up time as
2439 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2440 it from "IPv6 and IPv4", which means that two separate sockets are being
2443 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2444 says whether AF_INET or AF_INET6 was passed as an argument.
2446 42. Exiwhat output was messed up when time zones were included in log
2449 43. Exiwhat now gives more information about the daemon's listening ports,
2450 and whether -tls-on-connect was used.
2452 44. The "port" option of the smtp transport is now expanded.
2454 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2455 silently ignored. Now an error message is written to the main and panic
2458 46. There's a new ACL modifier called "logwrite" which writes to a log file
2459 as soon as it is encountered.
2461 47. Added $local_user_uid and $local_user_gid at routing time.
2463 48. Exim crashed when trying to verify a sender address that was being
2466 49. Exim was recognizing only a space character after ".include". It now also
2467 recognizes a tab character.
2469 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2470 extracting the relevant information from the specification. The man page no
2471 longer contains scrambled data for the -d option, and I've added a section
2472 at the front about calling Exim under different names.
2474 51. Added "extra_headers" argument to the "mail" command in filter files.
2476 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2479 53. Installed eximstats 1.29.
2481 54. Added transport_filter_timeout as a generic transport option.
2483 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2484 Cc: header lines. This was required by RFC 822, but it not required by RFC
2487 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2488 incoming messages that did not have them. Now it does so only if the
2489 message originates locally, that is, if there is no associated remote host
2490 address. When Resent- header lines are present, this applies to the Resent-
2491 lines rather than the non-Resent- lines.
2493 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2494 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2496 58. Messages for configuration errors now include the name of the main
2497 configuration file - useful now that there may be more than one file in a
2498 list (.included file names were always shown).
2500 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2501 for those rare installations that do not start the daemon as root or run it
2502 setuid root. I've cut out the call to initgroups() if the daemon is not
2505 60. The Exim user and group can now be bound into the binary as text strings
2506 that are looked up at the start of Exim's processing.
2508 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2510 62. Added $mailstore_basename variable.
2512 63. Installed patch to sieve.c from Michael Haardt.
2514 64. When Exim failed to open the panic log after failing to open the main log,
2515 the original message it was trying to log was written to stderr and debug
2516 output, but if they were not available (the usual case in production), it
2517 was lost. Now it is written to syslog before the two lines that record the
2518 failures to open the logs.
2520 65. Users' Exim filters run in subprocesses under the user's uid. It is
2521 possible for a "deliver" command or an alias in a "personal" command to
2522 provoke an address rewrite. If logging of address rewriting is configured,
2523 this fails because the process is not running as root or exim. There may be
2524 a better way of dealing with this, but for the moment (because 4.30 needs
2525 to be released), I have disabled address rewrite logging when running a
2526 filter in a non-root, non-exim process.
2532 1. The buildconfig auxiliary program wasn't quoting the value set for
2533 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2534 not defined. This bug was masked in 4.22 by the effect that was fixed in
2537 2. Some messages that were rejected after a message id was allocated were
2538 shown as "incomplete" by exigrep. It no longer does this for messages that
2539 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2541 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2542 have allow_domain_literals set, the ID did not get logged in the <= line.
2543 Domain literals are now always recognized in Message-ID: header lines.
2545 4. The first argument for a ${extract expansion item is the key name or field
2546 number. Leading and trailing spaces in this item were not being ignored,
2547 causing some misleading effects.
2549 5. When deliver_drop_privilege was set, single queue runner processes started
2550 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2551 same command in the process it spins off) were not dropping privilege.
2553 6. When the daemon running as "exim" started a queue runner, it always
2554 re-executed Exim in the spun-off process. This is a waste of effort when
2555 deliver_drop_privilege is set. The new process now just calls the
2556 queue-runner function directly.
2562 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2565 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2566 ignored. Though the use of -oP was forcing the writing of a pid file, it
2567 was always written to the default place.
2569 3. If the message "no IP address found for host xxxx" is generated during
2570 incoming verification, it is now followed by identification of the incoming
2571 connection (so you can more easily find what provoked it).
2573 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2575 5. Added some features to "harden" Exim a bit more against certain attacks:
2577 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2578 be put in Local/Makefile. This is like the never_users runtime option,
2579 but it cannot be overridden. The default setting is "root".
2581 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2582 prefix string with which any file named in a -C command line option
2585 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2586 is retained for -C and -D only if the caller of Exim is root. Without
2587 it, the exim user may also use -C and -D and retain privilege.
2589 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2590 command line option is disabled.
2592 6. Macro names set by the -D option must start with an upper case letter, just
2593 like macro names defined in the configuration file.
2595 7. Added "dereference=" facility to LDAP.
2597 8. Two instances of the typo "uknown" in the source files are fixed.
2599 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2600 the Configure-Makefile script screwed up while processing it.
2602 10. Incorporated PCRE 4.4.
2604 11. The SMTP synchronization check was not operating right at the start of an
2605 SMTP session. For example, it could not catch a HELO sent before the client
2606 waited for the greeting. There is now a check for outstanding input at the
2607 point when the greeting is written. Because of the duplex, asynchronous
2608 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2609 way, but not yet received, when the check is performed.
2611 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2612 on TCP/IP sockets, because this apparently causes some broken clients to
2615 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2616 unchanged) from the Cygwin maintainer.
2618 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2619 was supported instead of testing for mbx. Effectively a typo.
2621 15. The spa authenticator server code was not checking that the input it
2622 received was valid base64.
2624 16. The debug output line for the "set" modifier in ACLs was not showing the
2625 name of the variable that was being set.
2627 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2629 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2630 lookup on its own. Something else was needed to trigger the lookup. For
2631 example, a match in host_lookup or the need for a host name in a host list.
2632 Now, if $sender_host_name is referenced and the host name has not yet been
2633 looked up, a lookup is performed. If the lookup fails, the variable remains
2634 empty, and $host_lookup_failed is set to "1".
2636 19. Added "eqi" as a case-independent comparison operator.
2638 20. The saslauthd authentication condition could segfault if neither service
2639 nor realm was specified.
2641 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2642 error message that was logged was misleading, and incoming SMTP
2643 connections were dropped. The message is now more accurate, and temporary
2644 errors are given to SMTP connections.
2646 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2647 (except RSET) with a 421 error, until QUIT is received. However, it was
2648 failing to send a response to QUIT.
2650 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2651 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2652 ACL is now run after the helo_try_verify_hosts code.
2654 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2655 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2656 case-independent, so other case variants are also recognized). Apparently
2657 some systems use these upper case variants.
2659 25. If more than two messages were waiting for the same host, and a transport
2660 filter was specified for the transport, Exim sent two messages over the
2661 same TCP/IP connection, and then failed with "socket operation on non-
2662 socket" when it tried to send the third.
2664 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2666 27. The extern definition of crypt16() in expand.c was not being excluded when
2667 the OS had its own crypt16() function.
2669 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2670 as a preferred synonym for return_size_limit, both as an option and as an
2673 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2675 30. Changed the default configuration ACL to relax the local part checking rule
2676 for addresses that are not in any local domains. For these addresses,
2677 slashes and pipe symbols are allowed within local parts, but the sequence
2678 /../ is explicitly forbidden.
2680 31. SPA server authentication was not clearing the challenge buffer before
2683 32. log_message in a "warn" ACL statement was writing to the reject log as
2684 well as to the main log, which contradicts the documentation and doesn't
2685 seem right (because no rejection is happening). So I have stopped it.
2687 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2688 However, I am unable to do any testing of this.
2690 34. Fixed an infelicity in the appendfile transport. When checking directories
2691 for a mailbox, to see if any needed to be created, it was accidentally
2692 using path names with one or more superfluous leading slashes; tracing
2693 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2695 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2696 contents are added to the log line that is written for every discarded
2697 recipient. (Previously a log_message setting was ignored.)
2699 36. The ${quote: operator now quotes the string if it is empty.
2701 37. The install script runs exim in order to find its version number. If for
2702 some reason other than non-existence or emptiness, which it checks, it
2703 could not run './exim', it was installing it with an empty version number,
2704 i.e. as "exim-". This error state is now caught, and the installation is
2707 38. An argument was missing from the function that creates an error message
2708 when Exim fails to connect to the socket for saslauthd authentication.
2709 This could cause Exim to crash, or give a corrupted message.
2711 39. Added isip, isip4, and isip6 to ${if conditions.
2713 40. The ACL variables $acl_xx are now saved with the message, and can be
2714 accessed later in routers, transports, and filters.
2716 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2717 strings in the file are not string-expanded.
2719 42. If a MAIL command specified a SIZE value that was too large to fit into an
2720 int variable, the check against message_size_limit failed. Such values are
2721 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2722 day this will have to be increased, but I don't think I want to be around
2723 when emails are that large.
2730 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2731 iconv() is not standard in FreeBSD.
2733 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2734 IPv6 enabled. The observed symptom was a segmentation fault on return from
2735 the function os_common_find_running_interfaces() in src/os.c.
2737 3. In the check_special_case() function in daemon.c I had used "errno" as an
2738 argument name, which causes warnings on some systems. This was basically a
2739 typo, since it was named "eno" in the comments!
2741 4. The code that waits for the clock to tick (at a resolution of some fraction
2742 of a second) so as to ensure message-id uniqueness was always waiting for
2743 at least one whole tick, when it could have waited for less. [This is
2744 almost certainly not relevant at current processor speeds, where it is
2745 unlikely to ever wait at all. But we try to future-proof.]
2747 5. The function that sleeps for a time interval that includes fractions of a
2748 second contained a race. It did not block SIGALRM between setting the
2749 timer, and suspending (a couple of lines later). If the interval was short
2750 and the sigsuspend() was delayed until after it had expired, the suspension
2751 never ended. On busy systems this could lead to processes getting stuck for
2754 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2755 process, before it forks any delivery processes. The open lookup caching
2756 mechanism meant that the open file or database connection was passed into
2757 the delivery process. The problem was that delivery processes always tidy
2758 up cached lookup data. This could cause a problem for the next delivery
2759 process started by the queue runner, because the external queue runner
2760 process does not know about the closure. So the next delivery process
2761 still has data in the lookup cache. In the case of a file lookup, there was
2762 no problem because closing a file descriptor in a subprocess doesn't affect
2763 the parent. However, if the lookup was caching a connection to a database,
2764 the connection was closed, and the second delivery process was likely to
2765 see errors such as "PGSQL: query failed: server closed the connection
2766 unexpectedly". The problem has been fixed by closing all cached lookups
2767 in a queue runner before running a delivery process.
2769 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2770 seem to have the "const" qualifier which it has on other OS. I've
2773 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2774 *of the same type* (client or server) with the same public name that an
2775 error should be diagnosed.
2777 9. When Exim looked up a host name for an IP address, but failed to find the
2778 original IP address when looking up the host name (a safety check), it
2779 output the message "<ip address> does not match any IP for NULL", which was
2780 confusing, to say the least. The bug was that the host name should have
2781 appeared instead of "NULL".
2783 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2784 user that is built into the binary, and the -C or -D options is used, root
2785 privilege is dropped before the configuration file is read. In addition,
2786 logging is switched to stderr instead of the normal log files. If the
2787 configuration then re-defines the Exim user, the unprivileged environment
2788 is probably not what is expected, so Exim logs a panic warning message (but
2791 However, if deliver_drop_privilege is set, the unprivileged state may well
2792 be exactly what is intended, so the warning has been cut out in that case,
2793 and Exim is allowed to try to write to its normal log files.
2799 1. smtp_return_error_details was not giving details for temporary sender
2800 or receiver verification errors.
2802 2. Diagnose a configuration error if two authenticators have the same public
2805 3. Exim used not to create the message log file for a message until the first
2806 delivery attempt. This could be confusing when incoming messages were held
2807 for policy or load reasons. The message log file is now created at the time
2808 the message is received, and an initial "Received" line is written to it.
2810 4. The automatically generated man page for command line options had a minor
2811 bug that caused no ill effects; however, a more serious problem was that
2812 the procedure for building the man page automatically didn't always
2813 operate. Consequently, release 4.20 contains an out-of-date version. This
2814 shouldn't happen again.
2816 5. When building Exim with embedded Perl support, the script that builds the
2817 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2818 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2820 6. The freeze_tell option was not being used for messages that were frozen on
2821 arrival, either by an ACL or by local_scan().
2823 7. Added the smtp_incomplete_transaction log selector.
2825 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2826 was accepting AUTH without a new EHLO.
2828 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2829 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2832 10. Logging of TCP/IP connections (when configured) now happens in the main
2833 daemon process instead of the child process, so that the TCP/IP connection
2834 count is more accurate (but it can never be perfect).
2836 11. The use of "drop" in a nested ACL was not being handled correctly in the
2837 outer ACL. Now, if condition failure induced by the nested "drop" causes
2838 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2839 or "require"), the connection is dropped.
2841 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2842 that yield "discard" can now be used with an "accept" or a "discard" verb,
2843 but an error is generated for any others (because I can't see a useful way
2844 to define what should happen).
2846 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2847 are now processed in the same way as lines in the Exim configuration file.
2848 In particular, continuation lines are supported.
2850 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2852 15. Added -ti meaning -t -i.
2854 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2855 domains, and warn by logging if others are found.
2857 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2858 returned from GIFCONF when Exim is trying to find the list of interfaces on
2859 a host. The code in os.c has been modified to copy each ifreq to an aligned
2860 structure in all cases.
2862 Also, in some cases, the returned ifreq's were being copied to a 'struct
2863 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2864 means the last couple of bytes of an IPv6 address could be chopped if the
2865 ifreq contained only a normal sockaddr (14 bytes storage).
2867 18. Named domain lists were not supported in the hosts_treat_as_local option.
2868 An entry such as +xxxx was not recognized, and was treated as a literal
2871 19. Ensure that header lines added by a DATA ACL are included in the reject log
2872 if the ACL subsequently rejects the message.
2874 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2875 MD5 (which is deprecated).
2877 21. When testing a filter file using -bf, Exim was writing a message when it
2878 took the sender from a "From " line in the message, but it was not doing so
2879 when it took $return_path from a Return-Path: header line. It now does.
2881 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2882 with a valid header line field name (a series of printing characters
2883 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2885 23. Changed "disc" in the source to "disk" to conform to the documentation and
2886 the book and for uniformity.
2888 24. Ignore Sendmail's -Ooption=value command line item.
2890 25. When execve() failed while trying to run a command in a pipe transport,
2891 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2892 could be confused with a return value of 69 from the command itself. This
2893 has been changed to 127, the value the shell returns if it is asked to run
2894 a non-existent command. The wording for the related log line suggests a
2895 non-existent command as the problem.
2897 26. If received_header_text expands to an empty string, do not add a Received:
2898 header line to the message. (Well, it adds a token one on the spool, but
2899 marks it "old" so that it doesn't get used or transmitted.)
2901 27. Installed eximstats 1.28 (addition of -nt option).
2903 28. There was no check for failure on the call to getsockname() in the daemon
2904 code. This can fail if there is a shortage of resources on the system, with
2905 ENOMEM, for example. A temporary error is now given on failure.
2907 29. Contrary to the C standard, it seems that in some environments, the
2908 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2909 program. Exim now does this explicitly; it affects the formatting of
2910 timestamps using strftime().
2912 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2913 complaints. I've now initialized all the variables, to avoid this.
2915 32. Header lines added by a system filter were not being "seen" during
2916 transport-time rewrites.
2918 33. The info_callback() function passed to OpenSSL is set up with type void
2919 (*)(SSL *, int, int), as described somewhere. However, when calling the
2920 function (actually a macro) that sets it up, the type void(*)() is
2921 expected. I've put in a cast to prevent warnings from picky compilers.
2923 34. If a DNS black list lookup found a CNAME record, but there were no A
2924 records associated with the domain it pointed at, Exim crashed.
2926 35. If a DNS black list lookup returned more than one A record, Exim ignored
2927 all but the first. It now scans all returned addresses if a particular IP
2928 value is being sought. In this situation, the contents of the
2929 $dnslist_value variable are a list of all the addresses, separated by a
2932 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2933 to accept a host name and all its aliases if the forward lookup for any of
2934 them yielded the IP address of the incoming connection. Now it accepts only
2935 those names whose forward lookup yields the correct IP address. Any other
2936 names are discarded. This closes a loophole whereby a rogue DNS
2937 administrator could create reverse DNS records to break through a
2938 wildcarded host restriction in an ACL.
2940 37. If a user filter or a system filter that ran in a subprocess used any of
2941 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2942 the wrong values were passed to the pipe command ($thisaddress had the
2943 value of $0, $0 had the value of $1, etc). This bug was introduced by
2944 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2946 38. Improved the line breaking for long SMTP error messages from ACLs.
2947 Previously, if there was no break point between 40 and 75 characters, Exim
2948 left the rest of the message alone. Two changes have been made: (a) I've
2949 reduced the minimum length to 35 characters; (b) if it can't find a break
2950 point between 35 and 75 characters, it looks ahead and uses the first one
2951 that it finds. This may give the occasional overlong line, but at least the
2952 remaining text gets split now.
2954 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2955 file descriptors might be low, and that setting 1000 would always raise it.
2956 It turns out that in some environments, the limit is already over 1000 and
2957 that lowering it causes trouble. So now Exim takes care not to decrease it.
2959 40. When delivering a message, the value of $return_path is set to $sender_
2960 address at the start of routing (routers may change the value). By an
2961 oversight, this default was not being set up when an address was tested by
2962 -bt or -bv, which affected the outcome if any router or filter referred to
2965 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2966 log only when "message" or "log_message" are set. However, if one of the
2967 conditions was an address verification, or a call to a nested ACL, the
2968 messages generated by the underlying test were being passed through. This
2969 no longer happens. The underlying message is available in $acl_verify_
2970 message for both "message" and "log_message" expansions, so it can be
2971 passed through if needed.
2973 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2974 new expansion $bh_ to give the encoded byte string without charset
2975 translation. Translation happens only if iconv() is available; HAVE_ICONV
2976 indicates this at build time. HEADERS_CHARSET gives the charset to
2977 translate to; headers_charset can change it in the configuration, and
2978 "headers charset" can change it in an individual filter file.
2980 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
2981 that creates RFC 2047 encoded "words" labels them as that charset instead
2982 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2983 expansion operator; (ii) when Exim creates a From: line for a local
2984 message; (iii) when a header line is rewritten to include a "phrase" part.
2986 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
2987 buggy, causing it to skip the first lines of messages whose message ID
2988 ended in 'D'. This would not have bitten before Exim release 4.14, because
2989 message IDs were unlikely to end in 'D' before then. The effect was to have
2990 incorrect size information for certain domains.
2992 45. #include "config.h" was missing at the start of the crypt16.c module. This
2993 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
2996 46. If there was a timeout during a "random" callout check, Exim treated it as
2997 a failure of the random address, and carried on sending RSET and the real
2998 address. If the delay was just some slowness somewhere, the response to the
2999 original RCPT would be taken as a response to RSET and so on, causing
3000 mayhem of various kinds.
3002 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3003 when I implemented it. It didn't allow for the fact that some option values
3004 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3005 the right test for positive values.
3007 48. Domain names in DNS records are case-independent. Exim always looks them up
3008 in lower case. Some resolvers return domain names in exactly the case they
3009 appear in the zone file, that is, they may contain uppercase letters. Not
3010 all resolvers do this - some return always lower case. Exim was treating a
3011 change of case by a resolver as a change of domain, similar to a widening
3012 of a domain abbreviation. This triggered its re-routing code and so it was
3013 trying to route what was effectively the same domain again. This normally
3014 caused routing to fail (because the router wouldn't handle the domain
3015 twice). Now Exim checks for this case specially, and just changes the
3016 casing of the domain that it ultimately uses when it transmits the message
3019 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3022 50. If a filter generated a file delivery with a non-absolute name (possible if
3023 no home directory exists for the router), the forbid_file option was not
3026 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3027 the existing equality matching.
3029 52. Exim was using ints instead of ino_t variables in some places where it was
3030 dealing with inode numbers.
3032 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3033 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3034 called TMPDIR, and if it finds it is different, it changes its value.
3036 54. The smtp_printf() function is now made available to local_scan() so
3037 additional output lines can be written before returning. There is also an
3038 smtp_fflush() function to enable the detection of a dropped connection.
3039 The variables smtp_input and smtp_batched_input are exported to
3042 55. Changed the default runtime configuration: the message "Unknown user"
3043 has been removed from the ACL, and instead placed on the localuser router,
3044 using the cannot_route_message feature. This means that any verification
3045 failures that generate their own messages won't get overridden. Similarly,
3046 the "Unrouteable address" message that was in the ACL for unverifiable
3047 relay addresses has also been removed.
3049 56. Added hosts_avoid_esmtp to the smtp transport.
3051 57. The exicyclog script was not checking for the esoteric option
3052 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3053 will work only if exicyclog is run under the appropriate euid.
3055 58. Following a discussion on the list, the rules by which Exim recognises line
3056 endings on incoming messages have been changed. The -dropcr and drop_cr
3057 options are now no-ops, retained only for backwards compatibility. The
3058 following line terminators are recognized: LF CRLF CR. However, special
3059 processing applies to CR:
3061 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3062 nor a local message in the state where . is a terminator.
3064 (ii) If a bare CR is encountered in a header line, an extra space is added
3065 after the line terminator so as not to end the header. The reasoning
3066 behind this is that bare CRs in header lines are most likely either
3067 to be mistakes, or people trying to play silly games.
3069 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3070 was being incorrectly given as 18 bytes larger than it should have been.
3071 This is a VOB (very old bug).
3073 60. This may never have affected anything current, but just in case it has:
3074 When the local host is found other than at the start of a list of hosts,
3075 the local host, those with the same MX, and any that follow, are discarded.
3076 When the list in question was part of a longer list of hosts, the following
3077 hosts (not currently being processed) were also being discarded. This no
3078 longer happens. I'm not sure if this situation could ever has previously
3081 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3084 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3085 header. This is something that is recommended in a new Internet Draft, and
3086 is something that is documented as being done by Sendmail. There are two
3087 possible values. For messages generated by the autoreply transport, Exim
3090 Auto-Submitted: auto-replied
3092 whereas for all other generated messages (e.g. bounces) it adds
3094 Auto-Submitted: auto-generated
3096 63. The "personal" condition in filters now includes a test for the
3097 Auto-Submitted: header. If it contains the string "auto-" the message it
3098 not considered personal.
3100 64. Added rcpt_include_affixes as a generic transport option.
3102 65. Added queue_only_override (default true).
3104 66. Added the syslog_duplication option.
3106 67. If what should have been the first header line of a message consisted of
3107 a space followed by a colon, Exim was mis-interpreting it as a header line.
3108 It isn't of course - it is syntactically invalid and should therefore be
3109 treated as the start of the message body. The misbehaviour could have
3110 caused a number of strange effects, including loss of data in subsequent
3111 header lines, and spool format errors.
3113 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3114 client host had authenticated. This control can now be exercised by an ACL
3115 for more flexibility.
3117 69. By default, callouts do not happen when testing with -bh. There is now a
3118 variant, -bhc, which does actually run the callout code, including
3119 consulting and updating the callout cache.
3121 70. Added support for saslauthd authentication, courtesy of Alexander
3124 71. If statvfs() failed on the spool or log directories while checking their
3125 size for availability, Exim confusingly gave the error "space shortage".
3126 Furthermore, in debugging mode it crashed with a floating point exception.
3127 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3128 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3129 serious problem, Exim now writes to the main and panic logs when this
3130 happens, with details of the failure. It then refuses to accept the
3131 incoming message, giving the message "spool directory problem" or "log
3132 directory problem" with a 421 code for SMTP messages.
3134 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3135 0, 1, and 2 exist, because some OS complain for execs without them (see
3136 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3137 descriptors. However, the code omitted to check that the open succeeded,
3138 causing mysterious errors if for some reason the permissions on /dev/null
3139 got screwed. Now Exim writes a message to the main and panic logs, and
3140 bombs out if it can't open /dev/null.
3142 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3143 interact so that it is all more flexible. It is supposed to remain
3144 backwards compatible. Also added extra_local_interfaces.
3146 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3147 to bomb out with an assertion failure - to the client this appears as a
3148 connection drop. This problem occurs in the part of the code that was taken
3149 from the Samba project. Fortunately, the assertion is in a very simple
3150 function, so I have fixed this by reproducing the function inline in the
3151 one place where it is called, and arranging for authentication to fail
3152 instead of killing the process with assert().
3154 75. The SPA client code was not working when the server requested OEM rather
3155 than Unicode encoding.
3157 76. Added code to make require_files with a specific uid setting more usable in
3158 the case where statting the file as root fails - usually a non-root-mounted
3159 NFS file system. When this happens and the failure is EACCES, Exim now
3160 forks a subprocess and does the per-uid checking as the relevant uid.
3162 77. Added process_log_path.
3164 78. If log_file_path was not explicitly set, a setting of check_log_space or
3165 check_log_inodes was ignored.
3167 79. If a space check for the spool or log partitions fails, the incident is now
3168 logged. Of course, in the latter case the data may get lost...
3170 80. Added the %p formatting code to string_format() so that it can be used to
3171 print addresses in debug_print(). Adjusted all the address printing in the
3172 debugging in store.c to use %p rather than %d.
3174 81. There was a concern that a line of code in smtp_in.c could overflow a
3175 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3176 initially expressed, the concern was not well-founded, because trailing
3177 spaces are removed early. However, if the trailing spaces were followed by
3178 a NULL, they did not get removed, so the overflow was possible. Two fixes
3181 (a) I re-wrote the offending code in a cleaner fashion.
3182 (b) If an incoming SMTP command contains a NULL character, it is rejected
3185 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3186 runs initgroups(), so that if the Exim user is in any additional groups,
3187 they will be used during message reception.
3193 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob