1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.202 2005/08/08 10:48:26 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Added the "success_on_redirect" address verification option. See
10 NewStuff for rationale and an example.
12 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
14 PH/02 Patch to exigrep to allow it to work on syslog lines.
16 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
17 fread() to skip over the body file's header line, because in Cygwin the
18 header line is locked and is inaccessible.
20 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
21 co-exist for some time) to make it clear that it is the Exim ID that is
22 referenced, not the Message-ID: header line.
24 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
25 string_format() function, because snprintf() does not exist on all
28 PH/06 The use of forbid_filter_existstest now also locks out the use of the
29 ${stat: expansion item.
31 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
32 protocol synchronization error", to keep the pedants happy.
34 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
35 well as for IRIX systems, when gcc is being used. See the host.c source
38 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
40 PH/10 Named domain lists were not working if used in a queue_smtp_domains
43 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
44 transport and to the smtp transport in LMTP mode.
46 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
48 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
49 run a filter in a subprocess. This could lead to confusion in subsequent
50 lookups in the parent process. There should also be a search_tidyup() at
51 the end of the subprocess.
53 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
54 only if the host matched helo_try_verify_hosts, which caused the
55 verification to occur when the EHLO/HELO command was issued. The ACL just
56 tested the remembered result. Now, if a previous verification attempt has
57 not happened, "verify = helo" does it there and then.
59 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
61 TK/03 Fix log output including CR from clamd.
63 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
64 exist provoked a memory error which could cause a segfault.
66 PH/15 Installed PCRE 6.2
68 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
70 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
71 of the problem. Specifically, suggested +O2 rather than +O1 for the
74 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
80 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
82 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
83 command itself ended in error, the underlying message about the transport
84 filter timeout was being overwritten with the pipe command error. Now the
85 underlying error message should be appended to the second error message.
87 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
89 PH/02 Reduce the amount of output that "make" produces by default. Full output
90 can still be requested.
92 PH/03 The warning log line about a condition test deferring for a "warn" verb
93 was being output only once per connection, rather than after each
94 occurrence (because it was using the same function as for successful
95 "warn" verbs). This seems wrong, so I have changed it.
97 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
98 it should not have, which might have caused a crash in the right
99 circumstances, but probably never did.
101 PH/04 Installed a modified version of Tony Finch's patch to make submission
102 mode fix the return path as well as the Sender: header line, and to
103 add a /name= option so that you can make the user's friendly name appear
106 TF/03 Added the control = fakedefer ACL modifier.
108 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
109 Mark Lowes for thorough testing.
111 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
113 TK/03 Merged latest SRS patch from Miles Wilton.
115 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
116 with the definition in sysexits.h (which is #included earlier).
117 Fortunately, Exim does not actually use EX_OK. The code used to try to
118 preserve the sysexits.h value, by assumimg that macro definitions were
119 scanned for macro replacements. I have been disabused of this notion,
120 so now the code just undefines EX_OK before #including unistd.h.
122 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
123 in the smtp transport. When a block could not be written in a single
124 write() function, the timeout was being re-applied to each part-write.
125 This seems wrong - if the receiver was accepting one byte at a time it
126 would take for ever. The timeout is now adjusted when this happens. It
127 doesn't have to be particularly precise.
129 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
130 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
132 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
133 without a domain if the check to <postmaster@domain> fails.
135 SC/01 Eximstats: added -xls and the ability to specify output files
136 (patch written by Frank Heydlauf).
138 SC/02 Eximstats: use FileHandles for outputing results.
140 SC/03 Eximstats: allow any combination of xls, txt, and html output.
142 SC/04 Eximstats: fixed display of large numbers with -nvr option
144 SC/05 Eximstats: fixed merging of reports with empty tables.
146 SC/06 Eximstats: added the -include_original_destination flag
148 SC/07 Eximstats: removed tabs and trailing whitespace.
150 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
152 TK/06 MBOX spool code: Add real "From " MBOX separator line
153 so the .eml file is really in mbox format (even though
154 most programs do not really care). Patch from Alex Miller.
156 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
157 The latter is generated from $received_to and is only set if the
158 message has one envelope recipient. SA can use these headers,
159 obviously out-of-the-box. Patch from Alex Miller.
161 PH/08 The ${def test on a variable was returning false if the variable's
162 value was "0", contrary to what the specification has always said!
163 The result should be true unless the variable is empty.
165 PH/09 The syntax error of a character other than { following "${if
166 def:variable_name" (after optional whitespace) was not being diagnosed.
167 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
168 accidental colon was present, for example, could give incorrect results.
170 PH/10 Tidied the code in a number of places where the st_size field of a stat()
171 result is used (not including appendfile, where other changes are about
174 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
175 This involved changing a lot of size variables from int to off_t. It
176 should work with maildirs and everything.
178 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
179 spamd dying while we are connected to it.
181 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
184 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
185 function for BDB 4.3.
187 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
189 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
190 the sources for PCRE have been re-arranged and more files are now
193 PH/15 The code I had for printing potentially long long variables in PH/11
194 above was not the best (it lost precision). The length of off_t variables
195 is now inspected at build time, and an appropriate printing format (%ld
196 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
197 to be "long long int" or "long int". This is needed for the internal
198 formatting function string_vformat().
200 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
201 the configuration file to be ":syslog", then the script "guesses" where
202 the logs files are, rather than using the compiled in default. In our
203 case the guess is not the same as the compiled default, so the script
204 suddenly stopped working when I started to use syslog. The patch checks
205 to see if log_file_path is "". If so, it attempts to read it from exim
206 with no configuration file to get the compiled in version, before it
207 falls back to the previous guessing code."
209 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
210 implementing BATV in an Exim configuration. See NewStuff for the gory
213 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
214 Makefile that are specific to HP-UX.
216 PH/18 If the "use_postmaster" option was set for a recipient callout together
217 with the "random" option, the postmaster address was used as the MAIL
218 FROM address for the random test, but not for the subsequent recipient
219 test. It is now used for both.
221 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
222 patch removes a few documentation additions to RFC 3028, because the
223 latest draft now contains them. It adds the new en;ascii-case comparator
224 and a new error check for 8bit text in MIME parts. Comparator and
225 require names are now matched exactly. I enabled the subaddress
226 extension, but it is not well tested yet (read: it works for me)."
228 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
229 rework some of the code of TK/09 above to avoid the hardwired use of
230 "%lld" and "long long". Replaced the call to snprintf() with a call to
233 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
234 records point to non-existent hosts", "retry timeout exceeded", and
235 "retry time not reached for any host after a long failure period".
237 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
238 experimental DomainKeys support:
240 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
241 (2) On an error such as an illegally used "control", the wrong name for
242 the control was given.
244 These problems did NOT occur unless DomainKeys support was compiled.
246 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
248 PH/24 Added ${if match_ip condition.
250 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
251 databases so that it will be absolutely obvious if a crash occurs in the
252 DB library. This is a regular occurrence (often caused by mis-matched
255 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
256 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
257 picked up on a user's system that detects such things. There doesn't seem
258 to be a gcc warning option for this - only an attribute that has to be
259 put on the function's prototype. It seems that in Fedora Core 4 they have
260 set this on a number of new functions. No doubt there will be more in due
263 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
264 specify a transport. However, if an address that was verified by such a
265 router was the subject of a callout, Exim crashed because it tried to
266 read the rcpt_include_affixes from the non-existent transport. Now it
267 just assumes that the setting of that option is false. This bug was
268 introduced by 4.51/PH/31.
270 PH/28 Changed -d+all to exclude +memory, because that information is very
271 rarely of interest, but it makes the output a lot bigger. People tend to
272 do -d+all out of habit.
274 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
275 code in os-type was giving problems when libc.so lives in lib64, like on
278 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
279 aren't the modern standard, and it seems that some systems' include files
280 don't always have them. Exim was already checking for some of the newer
281 ones like T_AAAA, and defining it itself. I've added checks for all the
282 record types that Exim uses.
284 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
285 not automatically generating a new one, as it is supposed to. This
286 prevented TLS from working. If the file did exist, but contained invalid
287 data, a new version was generated, as expected. It was only the case of a
288 non-existent file that was broken.
290 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
291 with a change in libdomainkeys > 0.64.
293 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
294 from DNS. If the selector record carries the flag, it now has
295 precedence over the domain-wide flag.
297 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
299 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
300 the use of a port name that isn't defined in /etc/services) occurred, the
301 message was deferred as in a normal delivery, and thus remained on the
302 spool, instead of being failed because of the mua_wrapper setting. This
303 is now fixed, and I tidied up some of the mua_wrapper messages at the
306 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
307 the messages in a hash of arrays rather than using individual hashes.
308 This is a bit cleaner and results in dramatic memory savings, albeit
309 at a slight CPU cost.
311 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
312 as requested by Marc Sherman.
314 SC/10 Eximstats: added histograms for user specified patterns as requested
317 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
319 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
320 fopen() in the content-scanning modules that did not already have it.
326 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
327 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
329 TK/02 Fix ACL "control" statement not being available in MIME ACL.
331 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
333 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
334 to test Sieve filters that use "vacation".
336 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
337 that changes the way the GnuTLS parameters are stored in the cache file.
338 The new format can be generated externally. For backward compatibility,
339 if the data in the cache doesn't make sense, Exim assumes it has read an
340 old-format file, and it generates new data and writes a new file. This
341 means that you can't go back to an older release without removing the
344 PH/03 A redirect router that has both "unseen" and "one_time" set does not
345 work if there are any delivery delays because "one_time" forces the
346 parent to be marked "delivered", so its unseen clone is never tried
347 again. For this reason, Exim now forbids the simultaneous setting of
350 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
351 redirected to themselves ("homonym" addresses). Read the long ChangeLog
352 entry if you want to know the details. The fix, however, neglected to
353 consider the case when local delivery batching is involved. The test for
354 "previously delivered" was not happening when checking to see if an
355 address could be batched with a previous (undelivered) one; under
356 certain circumstances this could lead to multiple deliveries to the same
359 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
360 in its include files, and this causes problems building Exim.
362 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
363 header_syntax) but Exim was just ignoring anything given after a slash.
364 In particular, this caused confusion with an attempt to use "verify =
365 reverse_host_lookup/defer_ok". An error is now given when options are
366 supplied for verify items that do not have them. (Maybe reverse_host_
367 lookup should have a defer_ok option, but that's a different point.)
369 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
370 defined by RFC 821) to 2048, because there were problems with some AUTH
371 commands, and RFC 1869 says the size should be increased for extended
372 SMTP commands that take arguments.
374 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
377 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
378 "unknown" error; now it says that the functionality isn't in the binary.
380 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
381 an address' error message when a string expansion fails (syntax or
382 whatever). Otherwise the password may appear in the log. Following change
383 PH/42 below, there is no longer a chance of it appearing in a bounce
386 PH/11 Installed exipick version 20050225.0 from John Jetmore.
388 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
389 of its addresses was ever tried. (Bugzilla bug #2.)
391 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
392 the result incorrectly in the debug output. (It correctly added a newline
393 to what was transported.)
395 TF/01 Added $received_time.
397 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
398 commented out examples of how to interface to a virus scanner and to
399 SpamAssassin. Also added commented examples of av_scanner and
400 spamd_address settings.
402 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
403 and controls are allowed in which ACLs. There were a couple of minor
404 errors. Some of the entries in the conditions table (which is a table of
405 where they are NOT allowed) were getting very unwieldy; rewrote them as a
406 negation of where the condition IS allowed.
408 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
410 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
411 header file does not have a version number, so I've had to invent a new
412 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
413 API. The code is untested by me (my Linux distribution still has 0.3.2 of
414 radiusclient), but it was contributed by a Radius user.
416 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
417 files or directories, for OpenSSL.
419 PH/19 When an Exim process that is running as root has to create an Exim log
420 file, it does so in a subprocess that runs as exim:exim so as to get the
421 ownership right at creation (otherwise, other Exim processes might see
422 the file with the wrong ownership). There was no test for failure of this
423 fork() call, which would lead to the process getting stuck as it waited
424 for a non-existent subprocess. Forks do occasionally fail when resources
425 run out. I reviewed all the other calls to fork(); they all seem to check
428 PH/20 When checking for unexpected SMTP input at connect time (before writing
429 the banner), Exim was not dealing correctly with a non-positive return
430 from the read() function. If the client had disconnected by this time,
431 the result was a log entry for a synchronization error with an empty
432 string after "input=" when read() returned zero. If read() returned -1
433 (an event I could not check), uninitialized data bytes were printed.
434 There were reports of junk text (parts of files, etc) appearing after
437 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
439 PH/22 Added support for macro redefinition, and (re)definition in between
440 driver and ACL definitions.
442 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
443 forgetting to use the resulting value; it was using the unexpanded value.
445 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
446 hadn't been configured. The fix is from Juergen Kreileder, who
447 understands it better than I do:
449 "Here's what I see happening with three configured cyrus_sasl
450 authenticators configured (plain, login, cram-md5):
452 On startup auth_cyrus_sasl_init() gets called for each of these.
453 This means three calls to sasl_listmech() without a specified mech_list.
454 => SASL tests which mechs of all available mechs actually work
455 => three warnings about OTP not working
456 => the returned list contains: plain, login, cram-md5, digest-md5, ...
458 With the patch, sasl_listmech() also gets called three times. But now
459 SASL's mech_list option is set to the server_mech specified in the the
460 authenticator. Or in other words, the answer from sasl_listmech()
461 gets limited to just the mech you're testing for (which is different
463 => the return list contains just 'plain' or 'login', 'cram-md5' or
464 nothing depending on the value of ob->server_mech.
466 I've just tested the patch: Authentication still works fine,
467 unavailable mechs specified in the exim configuration are still
468 caught, and the auth.log warnings about OTP are gone."
470 PH/25 When debugging is enabled, the contents of the command line are added
471 to the debugging output, even when log_selector=+arguments is not
474 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
475 answer is "GNU", and only if the return is "GNU/something" is the answer
478 PH/27 $acl_verify_message is now set immediately after the failure of a
479 verification in an ACL, and so is available in subsequent modifiers. In
480 particular, the message can be preserved by coding like this:
482 warn !verify = sender
483 set acl_m0 = $acl_verify_message
485 Previously, $acl_verify_message was set only while expanding "message"
486 and "log_message" when a very denied access.
488 PH/28 Modified OS/os.c-Linux with
490 -#ifndef OS_LOAD_AVERAGE
491 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
493 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
494 nomenclature these days.)
496 PH/29 Installed patch from the Sieve maintainer that adds the options
497 sieve_useraddress and sieve_subaddress to the redirect router.
499 PH/30 In these circumstances:
500 . Two addresses routed to the same list of hosts;
501 . First host does not offer TLS;
502 . First host accepts first address;
503 . First host gives temporary error to second address;
504 . Second host offers TLS and a TLS session is established;
505 . Second host accepts second address.
506 Exim incorrectly logged both deliveries with the TLS parameters (cipher
507 and peerdn, if requested) that were in fact used only for the second
510 PH/31 When doing a callout as part of verifying an address, Exim was not paying
511 attention to any local part prefix or suffix that was matched by the
512 router that accepted the address. It now behaves in the same way as it
513 does for delivery: the affixes are removed from the local part unless
514 rcpt_include_affixes is set on the transport.
516 PH/32 Add the sender address, as F=<...>, to the log line when logging a
517 timeout during the DATA phase of an incoming message.
519 PH/33 Sieve envelope tests were broken for match types other than :is. I have
520 applied a patch sanctioned by the Sieve maintainer.
522 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
523 the uid or gid is negative. A case of a negative gid caused this to be
524 noticed. The fix allows for either to be negative.
526 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
527 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
530 PH/36 The change PH/12 above was broken. Fixed it.
532 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
533 the grounds that routing the same address twice would always produce the
534 same answer. This might have been true once, but it is certainly no
535 longer true now. Routing a child address may depend on the previous
536 routing that produced that child. Some complicated redirection strategies
537 went wrong when messages had multiple recipients, and made Exim's
538 behaviour dependent on the order in which the addresses were given.
540 I have moved the duplicate checking until after the routing is complete.
541 Exim scans the addresses that are assigned to local and remote
542 transports, and removes any duplicates. This means that more work will be
543 done, as duplicates will always all be routed, but duplicates are
544 presumably rare, so I don't expect this is of any significance.
546 For deliveries to pipes, files, and autoreplies, the duplicate checking
547 still happens during the routing process, since they are not going to be
550 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
551 It corrects a timeout issue with spamd. This is Ian's comment: "The
552 background is that sometimes spamd either never reads data from a
553 connection it has accepted, or it never writes response data. The exiscan
554 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
555 blindly assumes that writes won't block so it may never time out."
557 PH/39 Allow G after quota size as well as K and M.
559 PH/40 The value set for $authenticated_id in an authenticator may not contain
560 binary zeroes or newlines because the value is written to log lines and
561 to spool files. There was no check on this. Now the value is run through
562 the string_printing() function so that such characters are converted to
563 printable escape sequences.
565 PH/41 $message_linecount is a new variable that contains the total number of
566 lines in the message. Compare $body_linecount, which is the count for the
569 PH/42 Exim no longer gives details of delivery errors for specific addresses in
570 bounce and delay warning messages, except in certain special cases, which
573 (a) An SMTP error message from a remote host;
574 (b) A message specified in a :fail: redirection;
575 (c) A message specified in a "fail" command in a system filter;
576 (d) A message specified in a FAIL return from the queryprogram router;
577 (e) A message specified by the cannot_route_message router option.
579 In these cases only, Exim does include the error details in bounce and
580 warning messages. There are also a few cases where bland messages such
581 as "unrouteable address" or "local delivery error" are given.
583 PH/43 $value is now also set for the "else" part of a ${run expansion.
585 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
586 being worked on, but at least Exim now implements the latest version to
589 PH/45 In a pipe transport, although a timeout while waiting for the pipe
590 process to complete was treated as a delivery failure, a timeout while
591 writing the message to the pipe was logged, but erroneously treated as a
592 successful delivery. Such timeouts include transport filter timeouts. For
593 consistency with the overall process timeout, these timeouts are now
594 treated as errors, giving rise to delivery failures by default. However,
595 there is now a new Boolean option for the pipe transport called
596 timeout_defer, which, if set TRUE, converts the failures into defers for
597 both kinds of timeout. A transport filter timeout is now identified in
600 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
601 systems where "make" and "gmake" are different, calling "gmake" at top
602 level broke things. I've arranged for the value of $(MAKE) to be passed
603 from the Makefile to this script so that it can call the same version of
607 A note about Exim versions 4.44 and 4.50
608 ----------------------------------------
610 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
611 changes of various kinds. As a consequence, a big documentation update was
612 needed. This delayed the release for rather longer than seemed good, especially
613 in the light of a couple of (minor) security issues. Therefore, the changes
614 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
615 release. So 4.44 and 4.50 are in effect two different branches that both start
618 I have left the 4.50 change log unchanged; it contains all the changes since
619 4.43. The change log for 4.44 is below; many of its items are identical to
620 those for 4.50. This seems to be the most sensible way to preserve the
621 historical information.
627 1. Minor wording change to the doc/README.SIEVE file.
629 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
630 computation of the current number of files was incorrect.
632 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
633 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
636 4. Give more explanation in the error message when the command for a transport
637 filter fails to execute.
639 5. There are several places where Exim runs a non-Exim command in a
640 subprocess. The SIGUSR1 signal should be disabled for these processes. This
641 was being done only for the command run by the queryprogram router. It is
642 now done for all such subprocesses. The other cases are: ${run, transport
643 filters, and the commands run by the lmtp and pipe transports.
645 6. Added CONFIGURE_GROUP build-time option.
647 7. Some older OS have a limit of 256 on the maximum number of file
648 descriptors. Exim was using setrlimit() to set 1000 as a large value
649 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
650 systems. I've change it so that if it can't get 1000, it tries for 256.
652 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
653 was an oversight, and furthermore, ever since the addition of extra
654 controls (e.g. 4.43/32), the checks on when to allow different forms of
655 "control" were broken. There should now be diagnostics for all cases when a
656 control that does not make sense is encountered.
658 9. Added the /retain_sender option to "control=submission".
660 10. $recipients is now available in the predata ACL (oversight).
662 11. Tidy the search cache before the fork to do a delivery from a message
663 received from the command line. Otherwise the child will trigger a lookup
664 failure and thereby defer the delivery if it tries to use (for example) a
665 cached ldap connection that the parent has called unbind on.
667 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
668 of $address_data from the recipient verification was clobbered by the
671 13. The value of address_data from a sender verification is now available in
672 $sender_address_data in subsequent conditions in the ACL statement.
674 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
676 15. Added a new option "connect=<time>" to callout options, to set a different
679 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
680 was its contents. (It was OK if the option was not defined at all.)
682 17. A "Completed" log line is now written for messages that are removed from
683 the spool by the -Mrm option.
685 18. New variables $sender_verify_failure and $recipient_verify_failure contain
686 information about exactly what failed.
688 19. Added -dd to debug only the daemon process.
690 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
691 handles timeouts, both on the server side and network timeouts. Renamed the
692 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
694 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
695 It is now set to "smtps".
697 22. $host_address is now set to the target address during the checking of
700 23. When checking ignore_target_hosts for an ipliteral router, no host name was
701 being passed; this would have caused $sender_host_name to have been used if
702 matching the list had actually called for a host name (not very likely,
703 since this list is usually IP addresses). A host name is now passed as
706 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
707 code that specifies a non-restarting handler (typically sigaction() in
708 modern systems) in an attempt to fix a rare and obscure crash bug.
710 25. Narrowed the window for a race in the daemon that could cause it to ignore
711 SIGCHLD signals. This is not a major problem, because they are used only to
712 wake it up if nothing else does.
714 26. A malformed maildirsize file could cause Exim to calculate negative values
715 for the mailbox size or file count. Odd effects could occur as a result.
716 The maildirsize information is now recalculated if the size or filecount
719 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
720 support for a long time. Removed HAVE_SYS_VFS_H.
722 28. Installed the latest version of exipick from John Jetmore.
724 29. In an address list, if the pattern was not a regular expression, an empty
725 subject address (from a bounce message) matched only if the pattern was an
726 empty string. Non-empty patterns were not even tested. This was the wrong
727 because it is perfectly reasonable to use an empty address as part of a
728 database query. An empty address is now tested by patterns that are
729 lookups. However, all the other forms of pattern expect the subject to
730 contain a local part and a domain, and therefore, for them, an empty
731 address still always fails if the pattern is not itself empty.
733 30. Exim went into a mad DNS loop when attempting to do a callout where the
734 host was specified on an smtp transport, and looking it up yielded more
737 31. Re-factored the code for checking spool and log partition space into a
738 function that finds that data and another that does the check. The former
739 is then used to implement four new variables: $spool_space, $log_space,
740 $spool_inodes, and $log_inodes.
742 32. The RFC2047 encoding function was originally intended for short strings
743 such as real names; it was not keeping to the 75-character limit for
744 encoded words that the RFC imposes. It now respects the limit, and
745 generates multiple encoded words if necessary. To be on the safe side, I
746 have increased the buffer size for the ${rfc2047: expansion operator from
749 33. It is now permitted to omit both strings after an "if" condition; if the
750 condition is true, the result is "true". As before, when the second string
751 is omitted, a false condition yields an empty string. This makes it less
752 cumbersome to write custom ACL and router conditions.
754 34. Failure to deliver a bounce message always caused it to be frozen, even if
755 there was an errors_to setting on the router. The errors_to setting is now
758 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
759 canonical form (fully expanded) before being placed in
760 $sender_host_address.
762 36. The table in the code that translates DNS record types into text (T_A to
763 "A" for instance) was missing entries for NS and CNAME. It is just possible
764 that this could have caused confusion if both these types were looked up
765 for the same domain, because the text type is used as part of Exim's
766 per-process caching. But the chance of anyone hitting this buglet seems
769 37. The dnsdb lookup has been extended in a number of ways.
771 (1) There is a new type, "zns", which walks up the domain tree until it
772 finds some nameserver records. It should be used with care.
774 (2) There is a new type, "mxh", which is like "mx" except that it returns
775 just the host names, not the priorities.
777 (3) It is now possible to give a list of domains (or IP addresses) to be
778 looked up. The behaviour when one of the lookups defers can be
779 controlled by a keyword.
781 (4) It is now possible to specify the separator character for use when
782 multiple records are returned.
784 38. The dnslists ACL condition has been extended: it is now possible to supply
785 a list of IP addresses and/or domains to be looked up in a particular DNS
788 39. Added log_selector=+queue_time_overall.
790 40. When running the queue in the test harness, wait just a tad after forking a
791 delivery process, to get repeatability of debugging output.
793 41. Include certificate and key file names in error message when GnuTLS fails
794 to set them up, because the GnuTLS error message doesn't include the name
795 of the failing file when there is a problem reading it.
797 42. Allow both -bf and -bF in the same test run.
799 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
801 44. The "Exiscan patch" is now merged into the mainline Exim source.
803 45. Sometimes the final signoff response after QUIT could fail to get
804 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
805 < 0 before doing a fflush(). This bug looks as though it goes back to the
806 introduction of TLS in release 3.20, but "sometimes" must have been rare
807 because the tests only now provoked it.
809 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
810 (this can affect the format of dates).
812 47. exim_tidydb, when checking for the continued existence of a message for
813 which it has found a message-specific retry record, was not finding
814 messages that were in split spool directories. Consequently, it was
815 deleting retry records that should have stayed in existence.
817 48. Steve fixed some bugs in eximstats.
819 49. The SPA authentication driver was not abandoning authentication and moving
820 on to the next authenticator when an expansion was forced to fail,
821 contradicting the general specification for all authenticators. Instead it
822 was generating a temporary error. It now behaves as specified.
824 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
825 (the order specifies the preference for clients). The order is now AES256,
826 AES128, 3DES, ARCFOUR128.
828 51. Small patch to Sieve code - explicitly set From: when generating an
831 52. Exim crashed if a remote delivery caused a very long error message to be
832 recorded - for instance if somebody sent an entire SpamAssassin report back
833 as a large number of 550 error lines. This bug was coincidentally fixed by
834 increasing the size of one of Exim's internal buffers (big_buffer) that
835 happened as part of the Exiscan merge. However, to be on the safe side, I
836 have made the code more robust (and fixed the comments that describe what
839 53. Now that there can be additional text after "Completed" in log lines (if
840 the queue_time_overall log selector is set), a one-byte patch to exigrep
841 was needed to allow it to recognize "Completed" as not the last thing in
844 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
845 patch that reportedly fixes this has been added. I am not expert enough to
846 create a test for it. This is what the patch creator wrote:
848 "I found a little strange behaviour of ldap code when working with
849 Windows 2003 AD Domain, where users was placed in more than one
850 Organization Units. When I tried to give exim partial DN, the exit code
851 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
852 But simultaneously result of request was absolutely normal ldap result,
853 so I produce this patch..."
855 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
856 REFERENCE, so I have modified the code to exclude the patch when that macro
859 55. Some experimental protocols are using DNS PTR records for new purposes. The
860 keys for these records are domain names, not reversed IP addresses. The
861 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
862 leaves it alone. Component reversal etc. now happens only for IP addresses.
865 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
867 57. Double the size of the debug message buffer (to 2048) so that more of very
868 long debug lines gets shown.
870 58. The exicyclog utility now does better if the number of log files to keep
871 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
873 59. Two changes related to the smtp_active_hostname option:
875 (1) $smtp_active_hostname is now available as a variable.
876 (2) The default for smtp_banner uses $smtp_active_hostname instead
877 of $primary_hostname.
879 60. The host_aton() function is supposed to be passed a string that is known
880 to be a valid IP address. However, in the case of IPv6 addresses, it was
881 not checking this. This is a hostage to fortune. Exim now panics and dies
882 if the condition is not met. A case was found where this could be provoked
883 from a dnsdb PTR lookup with an IPv6 address that had more than 8
884 components; fortuitously, this particular loophole had already been fixed
885 by change 4.50/55 above.
887 If there are any other similar loopholes, the new check in host_aton()
888 itself should stop them being exploited. The report I received stated that
889 data on the command line could provoke the exploit when Exim was running as
890 exim, but did not say which command line option was involved. All I could
891 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
895 61. There was a buffer overflow vulnerability in the SPA authentication code
896 (which came originally from the Samba project). I have added a test to the
897 spa_base64_to_bits() function which I hope fixes it.
900 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
901 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
903 63. The daemon start-up calls getloadavg() while still root for those OS that
904 need the first call to be done as root, but it missed one case: when
905 deliver_queue_load_max is set with deliver_drop_privilege. This is
906 necessary for the benefit of the queue runner, because there is no re-exec
907 when deliver_drop_privilege is set.
909 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
912 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
913 was in use, was not putting the data itself into the right store pool;
914 consequently, it could be overwritten for a subsequent message in the same
915 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
918 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
920 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
921 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
922 the function were treating the return as a boolean value, which happened to
923 work because 0=false and not-0=true, but is not correct code.
925 68. The host_aton() function was not handling scoped IPv6 addresses (those
926 with, for example, "%eth0" on the end) correctly.
928 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
929 negated items (that is, ~something) in unsigned ints. Some compilers
930 apparently mutter when there is no cast.
932 70. If an address verification called from an ACL failed, and did not produce a
933 user-specific message (i.e. there was only a "system" message), nothing was
934 put in $acl_verify_message. In this situation, it now puts the system
937 71. Change 4.23/11 added synchronization checking at the start of an SMTP
938 session; change 4.31/43 added the unwanted input to the log line - except
939 that it did not do this in the start of session case. It now does.
941 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
942 This is wrong and can cause the other end to generate a synchronization
943 error if it is another Exim or anything else that does the synchronization
944 check. A QUIT command is no longer sent after a timeout.
946 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
949 74. The defer_ok option of callout verification was not working if it was used
950 when verifying addresses in header lines, that is, for this case:
952 verify = header_sender/callout=defer_ok
954 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
955 those file descriptors could be used for SMTP connections. If anything
956 wrote to stderr (the example that came up was "warn" in embedded Perl), it
957 could be sent to the SMTP client, causing chaos. The daemon now opens
958 stdin, stdout, and stderr to /dev/null when it puts itself into the
961 76. Arrange for output from Perl's "warn" command to be written to Exim's main
962 log by default. The user can override this with suitable Perl magic.
964 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
965 the log message when discard triggers, was not working for the DATA ACL or
966 for the non-SMTP ACL.
968 78. Error message wording change in sieve.c.
970 79. If smtp_accept_max_per_host was set, the number of connections could be
971 restricted to fewer than expected, because the daemon was trying to set up
972 a new connection before checking whether the processes handling previous
973 connections had finished. The check for completed processes is now done
974 earlier. On busy systems, this bug wouldn't be noticed because something
975 else would have woken the daemon, and it would have reaped the completed
978 80. If a message was submitted locally by a user whose login name contained one
979 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
980 It caused a spool format error. I have fixed the spool reading code. A
981 related problem was that the "from" clause in the Received: line became
982 illegal because of the space(s). It is now covered by ${quote_local_part.
984 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
987 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
989 83. Patch from Sieve maintainer to fix unterminated string problem in
992 84. Some minor changes to the Linux configuration files to help with other
993 OS variants using glibc.
995 85. One more patch for Sieve to update vacation handling to latest spec.
998 ----------------------------------------------------
999 See the note above about the 4.44 and 4.50 releases.
1000 ----------------------------------------------------
1006 1. Change 4.43/35 introduced a bug that caused file counts to be
1007 incorrectly computed when quota_filecount was set in an appendfile
1010 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1011 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1014 3. Give more explanation in the error message when the command for a transport
1015 filter fails to execute.
1017 4. There are several places where Exim runs a non-Exim command in a
1018 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1019 was being done only for the command run by the queryprogram router. It is
1020 now done for all such subprocesses. The other cases are: ${run, transport
1021 filters, and the commands run by the lmtp and pipe transports.
1023 5. Some older OS have a limit of 256 on the maximum number of file
1024 descriptors. Exim was using setrlimit() to set 1000 as a large value
1025 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1026 systems. I've change it so that if it can't get 1000, it tries for 256.
1028 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1029 was an oversight, and furthermore, ever since the addition of extra
1030 controls (e.g. 4.43/32), the checks on when to allow different forms of
1031 "control" were broken. There should now be diagnostics for all cases when a
1032 control that does not make sense is encountered.
1034 7. $recipients is now available in the predata ACL (oversight).
1036 8. Tidy the search cache before the fork to do a delivery from a message
1037 received from the command line. Otherwise the child will trigger a lookup
1038 failure and thereby defer the delivery if it tries to use (for example) a
1039 cached ldap connection that the parent has called unbind on.
1041 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1042 of $address_data from the recipient verification was clobbered by the
1043 sender verification.
1045 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1046 was its contents. (It was OK if the option was not defined at all.)
1048 11. A "Completed" log line is now written for messages that are removed from
1049 the spool by the -Mrm option.
1051 12. $host_address is now set to the target address during the checking of
1052 ignore_target_hosts.
1054 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1055 being passed; this would have caused $sender_host_name to have been used if
1056 matching the list had actually called for a host name (not very likely,
1057 since this list is usually IP addresses). A host name is now passed as
1060 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1061 code that specifies a non-restarting handler (typically sigaction() in
1062 modern systems) in an attempt to fix a rare and obscure crash bug.
1064 15. Narrowed the window for a race in the daemon that could cause it to ignore
1065 SIGCHLD signals. This is not a major problem, because they are used only to
1066 wake it up if nothing else does.
1068 16. A malformed maildirsize file could cause Exim to calculate negative values
1069 for the mailbox size or file count. Odd effects could occur as a result.
1070 The maildirsize information is now recalculated if the size or filecount
1073 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1074 support for a long time. Removed HAVE_SYS_VFS_H.
1076 18. Updated exipick to current release from John Jetmore.
1078 19. Allow an empty sender to be matched against a lookup in an address list.
1079 Previously the only cases considered were a regular expression, or an
1082 20. Exim went into a mad DNS lookup loop when doing a callout where the
1083 host was specified on the transport, if the DNS lookup yielded more than
1086 21. The RFC2047 encoding function was originally intended for short strings
1087 such as real names; it was not keeping to the 75-character limit for
1088 encoded words that the RFC imposes. It now respects the limit, and
1089 generates multiple encoded words if necessary. To be on the safe side, I
1090 have increased the buffer size for the ${rfc2047: expansion operator from
1093 22. Failure to deliver a bounce message always caused it to be frozen, even if
1094 there was an errors_to setting on the router. The errors_to setting is now
1097 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1098 canonical form (fully expanded) before being placed in
1099 $sender_host_address.
1101 24. Updated eximstats to version 1.33
1103 25. Include certificate and key file names in error message when GnuTLS fails
1104 to set them up, because the GnuTLS error message doesn't include the name
1105 of the failing file when there is a problem reading it.
1107 26. Expand error message when OpenSSL has problems setting up cert/key files.
1110 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1111 (this can affect the format of dates).
1113 28. exim_tidydb, when checking for the continued existence of a message for
1114 which it has found a message-specific retry record, was not finding
1115 messages that were in split spool directories. Consequently, it was
1116 deleting retry records that should have stayed in existence.
1118 29. eximstats updated to version 1.35
1119 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1120 1.35 - bugfix such that pie charts by volume are generated correctly
1122 30. The SPA authentication driver was not abandoning authentication and moving
1123 on to the next authenticator when an expansion was forced to fail,
1124 contradicting the general specification for all authenticators. Instead it
1125 was generating a temporary error. It now behaves as specified.
1127 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1128 (the order specifies the preference for clients). The order is now AES256,
1129 AES128, 3DES, ARCFOUR128.
1131 31. Small patch to Sieve code - explicitly set From: when generating an
1134 32. Exim crashed if a remote delivery caused a very long error message to be
1135 recorded - for instance if somebody sent an entire SpamAssassin report back
1136 as a large number of 550 error lines. This bug was coincidentally fixed by
1137 increasing the size of one of Exim's internal buffers (big_buffer) that
1138 happened as part of the Exiscan merge. However, to be on the safe side, I
1139 have made the code more robust (and fixed the comments that describe what
1142 33. Some experimental protocols are using DNS PTR records for new purposes. The
1143 keys for these records are domain names, not reversed IP addresses. The
1144 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1145 leaves it alone. Component reversal etc. now happens only for IP addresses.
1148 34. The host_aton() function is supposed to be passed a string that is known
1149 to be a valid IP address. However, in the case of IPv6 addresses, it was
1150 not checking this. This is a hostage to fortune. Exim now panics and dies
1151 if the condition is not met. A case was found where this could be provoked
1152 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1153 components; fortuitously, this particular loophole had already been fixed
1154 by change 4.50/55 or 4.44/33 above.
1156 If there are any other similar loopholes, the new check in host_aton()
1157 itself should stop them being exploited. The report I received stated that
1158 data on the command line could provoke the exploit when Exim was running as
1159 exim, but did not say which command line option was involved. All I could
1160 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1161 running as the user.
1164 35. There was a buffer overflow vulnerability in the SPA authentication code
1165 (which came originally from the Samba project). I have added a test to the
1166 spa_base64_to_bits() function which I hope fixes it.
1169 36. The daemon start-up calls getloadavg() while still root for those OS that
1170 need the first call to be done as root, but it missed one case: when
1171 deliver_queue_load_max is set with deliver_drop_privilege. This is
1172 necessary for the benefit of the queue runner, because there is no re-exec
1173 when deliver_drop_privilege is set.
1175 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1176 was in use, was not putting the data itself into the right store pool;
1177 consequently, it could be overwritten for a subsequent message in the same
1178 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1181 38. Sometimes the final signoff response after QUIT could fail to get
1182 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1183 < 0 before doing a fflush(). This bug looks as though it goes back to the
1184 introduction of TLS in release 3.20, but "sometimes" must have been rare
1185 because the tests only now provoked it.
1191 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1192 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1193 Now it returns an integer. A number of calls were still expecting a T/F
1194 return. Fortuitously, in all cases, the tests worked in OK situations,
1195 which is the norm. However, things would have gone wrong on any write
1196 failures on the smtp file descriptor. This function is used when sending
1197 messages over SMTP and also when doing verify callouts.
1199 2. When Exim is called to do synchronous delivery of a locally submitted
1200 message (the -odf or -odi options), it no longer closes stderr before doing
1203 3. Implemented the mua_wrapper option.
1205 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1207 5. Implemented the functions header_remove(), header_testname(),
1208 header_add_at_position(), and receive_remove_recipient(), and exported them
1211 6. If an ACL "warn" statement specified the addition of headers, Exim already
1212 inserted X-ACL-Warn: at the start if there was no header name. However, it
1213 was not making this test for the second and subsequent header lines if
1214 there were newlines in the string. This meant that an invalid header could
1215 be inserted if Exim was badly configured.
1217 7. Allow an ACL "warn" statement to add header lines at the start or after all
1218 the Received: headers, as well as at the end.
1220 8. Added the rcpt_4xx retry error code.
1222 9. Added postmaster_mailfrom=xxx to callout verification option.
1224 10. Added mailfrom=xxxx to the callout verification option, for verify=
1227 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1228 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1230 12. Inserted some casts to stop certain compilers warning when using pointer
1231 differences as field lengths or precisions in printf-type calls (mostly
1232 affecting debugging statements).
1234 13. Added optional readline() support for -be (dynamically loaded).
1236 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1237 same clock tick as a message's arrival, so that its received time was the
1238 same as the "first fail" time on the retry record, and that message
1239 remained on the queue past the ultimate address timeout, every queue runner
1240 would try a delivery (because it was past the ultimate address timeout) but
1241 after another failure, the ultimate address timeout, which should have then
1242 bounced the address, did not kick in. This was a "< instead of <=" error;
1243 in most cases the first failure would have been in the next clock tick
1244 after the received time, and all would be well.
1246 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1247 being recognized when the domain list was tested by the match_domain
1248 condition in an expansion string.
1250 16. Added the ${str2b64: operator.
1252 17. Exim was always calling setrlimit() to set a large limit for the number of
1253 processes, without checking whether the existing limit was already
1254 adequate. (It did check for the limit on file descriptors.) Furthermore,
1255 errors from getrlimit() and setrlimit() were being ignored. Now they are
1256 logged to the main and panic logs, but Exim does carry on, to try to do its
1257 job under whatever limits there are.
1259 18. Imported PCRE 5.0.
1261 19. Trivial typo in log message " temporarily refused connection" (the leading
1264 20. If the log selector return_path_on_delivery was set and an address was
1265 redirected to /dev/null, the delivery process crashed because it assumed
1266 that a return path would always be set for a "successful" delivery. In this
1267 case, the whole delivery is bypassed as an optimization, and therefore no
1270 21. Internal re-arrangement: the function for sending a challenge and reading
1271 a response while authentication was assuming a zero-terminated challenge
1272 string. It's now changed to take a pointer and a length, to allow for
1273 binary data in such strings.
1275 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1277 23. Exim was not respecting finduser_retries when seeking the login of the
1278 uid under which it was called; it was always trying 10 times. (The default
1279 setting of finduser_retries is zero.) Also, it was sleeping after the final
1280 failure, which is pointless.
1282 24. Implemented tls_on_connect_ports.
1284 25. Implemented acl_smtp_predata.
1286 26. If the domain in control=submission is set empty, Exim assumes that the
1287 authenticated id is a complete email address when it generates From: or
1288 Sender: header lines.
1290 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1291 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1292 chown and chgrp in /bin and hostname in /usr/bin.
1294 28. Exim was keeping the "process log" file open after each use, just as it
1295 does for the main log. This opens the possibility of it remaining open for
1296 long periods when the USR1 signal hits a daemon. Occasional processlog
1297 errors were reported, that could have been caused by this. Anyway, it seems
1298 much more sensible not to leave this file open at all, so that is what now
1301 29. The long-running daemon process does not normally write to the log once it
1302 has entered its main loop, and it closes the log before doing so. This is
1303 so that log files can straightforwardly be renamed and moved. However,
1304 there are a couple of unusual error situations where the daemon does write
1305 log entries, and I had neglected to close the log afterwards.
1307 30. The text of an SMTP error response that was received during a remote
1308 delivery was being truncated at 512 bytes. This is too short for some of
1309 the long messages that one sometimes sees. I've increased the limit to
1312 31. It is now possible to make retry rules that apply only when a message has a
1313 specific sender, in particular, an empty sender.
1315 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1316 it possible to be selective about when SMTP synchronization is enforced.
1318 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1320 32. Implemented hosts_connection_nolog.
1322 33. Added an ACL for QUIT.
1324 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1327 35. Added mailbox_size and mailbox_filecount to appendfile.
1329 36. Added control = no_multiline_responses to ACLs.
1331 37. There was a bug in the logic of the code that waits for the clock to tick
1332 in the case where the clock went backwards by a substantial amount such
1333 that the microsecond fraction of "now" was more than the microsecond
1334 fraction of "then" (but the whole seconds number was less).
1336 38. Added support for the libradius Radius client library this is found on
1337 FreeBSD (previously only the radiusclient library was supported).
1343 1. When certain lookups returned multiple values in the form name=value, the
1344 quoting of the values was not always being done properly. Specifically:
1345 (a) If the value started with a double quote, but contained no whitespace,
1347 (b) If the value contained whitespace other than a space character (i.e.
1348 tabs or newlines or carriage returns) it was not quoted.
1349 This fix has been applied to the mysql and pgsql lookups by writing a
1350 separate quoting function and calling it from the lookup code. The fix
1351 should probably also be applied to nisplus, ibase and oracle lookups, but
1352 since I cannot test any of those, I have not disturbed their existing code.
1354 2. A hit in the callout cache for a specific address caused a log line with no
1355 reason for rejecting RCPT. Now it says "Previous (cached) callout
1356 verification failure".
1358 3. There was an off-by-one bug in the queryprogram router. An over-long
1359 return line was truncated at 256 instead of 255 characters, thereby
1360 overflowing its buffer with the terminating zero. As well as fixing this, I
1361 have increased the buffer size to 1024 (and made a note to document this).
1363 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1364 when Exim is waiting for an SMTP response from a remote server, Exim
1365 restarts its select() call on the socket, thereby resetting its timeout.
1366 This is not a problem when such interrupts are rare. Somebody set up a cron
1367 job to run exiwhat every 2 minutes, which is less than the normal select()
1368 timeout (5 or 10 minutes). This meant that the select() timeout never
1369 kicked in because it was always reset. I have fixed this by comparing the
1370 time when an interrupt arrives with the time at the start of the first call
1371 to select(). If more time than the timeout has elapsed, the interrupt is
1372 treated as a timeout.
1374 5. Some internal re-factoring in preparation for the addition of Sieve
1375 extensions (by MH). In particular, the "personal" test is moved to a
1376 separate function, and given an option for scanning Cc: and Bcc: (which is
1377 not set for Exim filters).
1379 6. When Exim created an email address using the login of the caller as the
1380 local part (e.g. when creating a From: or Sender: header line), it was not
1381 quoting the local part when it contained special characters such as @.
1383 7. Installed new OpenBSD configuration files.
1385 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1386 try to make them clearer.
1388 9. Callout options, other than the timeout value, were being ignored when
1389 verifying sender addresses in header lines. For example, when using
1391 verify = header_sender/callout=no_cache
1393 the cache was (incorrectly) being used.
1395 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1396 only the Cygwin environment.
1398 11. When return_path_on_delivery was set as a log selector, if different remote
1399 addresses in the same message used different return paths and parallel
1400 remote delivery occurred, the wrong values would sometimes be logged.
1401 (Whenever a remote delivery process finished, the return path value from
1402 the most recently started remote delivery process was logged.)
1404 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1405 lines when AUTH and/or TLS are in use. This is the "received protocol"
1406 field. Exim used to use "asmtp" for authenticated SMTP, without any
1407 indication (in the protocol name) for TLS use. Now it follows the RFC and
1408 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1409 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1410 names appear in log lines as well as in Received: header lines.
1412 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1413 extensions, and comparison tests, and to fix some bugs.
1415 14. Changes to the "personal" filter test:
1417 (1) The test was buggy in that it was just doing the equivalent of
1418 "contains" tests on header lines. For example, if a user's address was
1419 anne@some.where, the "personal" test would incorrectly be true for
1421 To: susanne@some.where
1423 This test is now done by extracting each address from the header in turn,
1424 and checking the entire address. Other tests that are part of "personal"
1425 are now done using regular expressions (for example, to check local parts
1426 of addresses in From: header lines).
1428 (2) The list of non-personal local parts in From: addresses has been
1429 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1430 taken from the Sieve specification recommendations.
1432 (3) If the message contains any header line starting with "List-" it is
1433 treated as non-personal.
1435 (4) The test for "circular" in the Subject: header line has been removed
1436 because it now seems ill-conceived.
1438 15. Minor typos in src/EDITME comments corrected.
1440 16. Installed latest exipick from John Jetmore.
1442 17. If headers_add on a router specified a text string that was too long for
1443 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1444 of string_sprintf() is now avoided.
1446 18. $message_body_size was not set (it was always zero) when running the DATA
1447 ACL and the local_scan() function.
1449 19. For the "mail" command in an Exim filter, no default was being set for
1450 the once_repeat time, causing a random time value to be used if "once" was
1451 specified. (If the value happened to be <= 0, no repeat happened.) The
1452 default is now 0s, meaning "never repeat". The "vacation" command was OK
1453 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1454 (I found it when inspecting the code).
1456 20. There is now an overall timeout for performing a callout verification. It
1457 defaults to 4 times the callout timeout, which applies to individual SMTP
1458 commands during the callout. The overall timeout applies when there is more
1459 than one host that can be tried. The timeout is checked before trying the
1460 next host. This prevents very long delays if there are a large number of
1461 hosts and all are timing out (e.g. when the network connections are timing
1462 out). The value of the overall timeout can be changed by specifying an
1463 additional sub-option for "callout", called "maxwait". For example:
1465 verify = sender/callout=5s,maxwait=20s
1467 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1468 to the end before writing, but this should make it even safer).
1470 22. Exim was forgetting that it had advertised PIPELINING for the second and
1471 subsequent messages on an SMTP connection. It was also not resetting its
1472 memory on STARTTLS and an internal HELO.
1474 23. When Exim logs an SMTP synchronization error within a session, it now
1475 records whether PIPELINING has been advertised or not.
1477 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1478 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1479 rather than long int.
1481 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1483 26. Added the never_mail option to autoreply.
1489 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1490 crash if the getsockname() call failed; this can happen if a connection is
1491 closed very soon after it is established. The problem was simply in the
1492 order in which certain operations were done, causing Exim to try to write
1493 to the SMTP stream before it had set up the file descriptor. The bug has
1494 been fixed by making things happen in the correct order.
1500 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1501 before the connection was closed, thus losing the rejection response.
1503 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1504 some early Solaris releases, but causes trouble in current releases where
1505 socklen_t is defined.
1507 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1510 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1513 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1514 that releasing the top of it at the end releases what was used for sub-
1515 expansions (unless the block got too big). However, discard this block if
1516 the first thing is a variable or header, so that we can use its block when
1517 it is dynamic (useful for very large $message_headers, for example).
1519 6. Lookups now cache *every* query, not just the most recent. A new, separate
1520 store pool is used for this. It can be recovered when all lookup caches are
1521 flushed. Lookups now release memory at the end of their result strings.
1522 This has involved some general refactoring of the lookup sources.
1524 7. Some code has been added to the store_xxx() functions to reduce the amount
1525 of flapping under certain conditions.
1527 8. log_incoming_interface used to affect only the <= reception log lines. Now
1528 it causes the local interface and port to be added to several more SMTP log
1529 lines, for example "SMTP connection from", and rejection lines.
1531 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1533 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1535 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1536 could be overwritten at the end of the current message (or the start of a
1537 new message if it was set in a HELO ACL). The value is now preserved for
1538 the duration of the SMTP connection.
1540 12. If a transport had a headers_rewrite setting, and a matching header line
1541 contained an unqualified address, that address was qualified, even if it
1542 did not match any rewriting rules. The underlying bug was that the values
1543 of the flags that permit the existence of unqualified sender and recipient
1544 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1545 non-local messages, and by -bnq for local messages) were not being
1546 preserved with the message after it was received.
1548 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1549 "next input=" as part of the text comprising the host identity instead of
1550 the correct text. The code was using the same buffer for two different
1551 strings. However, depending on which order the printing function evaluated
1552 its arguments, the bug did not always show up. Under Linux, for example, my
1553 test suite worked just fine.
1555 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1556 doesn't work with some older versions of Perl. It has been changed to "my",
1557 which in any case is probably the better facility to use.
1559 15. A really picky compiler found some instances of statements for creating
1560 error messages that either had too many or two few arguments for the format
1563 16. The size of the buffer for calls to the DNS resolver has been increased
1564 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1565 for addresses that have a lot of PTR records. This alleviates a problem; it
1566 does not fully solve it.
1568 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1569 buffer now truncates the list and logs the incident, which is the same
1570 action as happens when Exim is looking up a host name and its aliases.
1571 Previously in this situation something unpredictable would happen;
1572 sometimes it was "internal error: store_reset failed".
1574 18. If a server dropped the connection unexpectedly when an Exim client was
1575 using GnuTLS and trying to read a response, the client delivery process
1576 crashed while trying to generate an error log message.
1578 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1579 string, for example:
1581 warn message = H1: something\nH2: something
1583 the text was added as a single header line from Exim's point of view
1584 though it ended up OK in the delivered message. However, searching for the
1585 second and subsequent header lines using $h_h2: did not work. This has been
1586 fixed. Similarly, if a system filter added multiple headers in this way,
1587 the routers could not see them.
1589 20. Expanded the error message when iplsearch is called with an invalid key to
1590 suggest using net-iplsearch in a host list.
1592 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1593 ACLs are no longer actually imposed (and a message to that effect is
1596 22. If a "gecos" field in a passwd entry contained escaped characters, in
1597 particular, if it contained a \" sequence, Exim got it wrong when building
1598 a From: or a Sender: header from that name. A second bug also caused
1599 incorrect handling when an unquoted " was present following a character
1600 that needed quoting.
1602 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1603 was not being matched caselessly.
1605 24. Arranged for all hyphens in the exim.8 source to be escaped with
1608 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1609 properly. Recipient callout cache records were still being keyed to include
1610 the sender, even when use_sender was set false. This led to far more
1611 callouts that were necessary. The sender is no longer included in the key
1612 when use_sender is false.
1614 26. Added "control = submission" modifier to ACLs.
1616 27. Added the ${base62d: operator to decode base 62 numbers.
1618 28. dnsdb lookups can now access SRV records.
1620 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1621 the configuration file.
1623 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1624 (-v) mode. This makes the output for a verbose queue run more intelligible.
1626 31. Added a use_postmaster feature to recipient callouts.
1628 32. Added the $body_zerocount variable, containing the number of binary zero
1629 bytes in the message body.
1631 33. The time of last modification of the "new" subdirectory is now used as the
1632 "mailbox time last read" when there is a quota error for a maildir
1635 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1637 35. Added +ignore_unknown as a special item in host lists.
1639 36. Code for decoding IPv6 addresses in host lists is now included, even if
1640 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1641 address was recognized as an IP address, but was then not correctly decoded
1642 into binary, causing unexpected and incorrect effects when compared with
1649 1. Very minor rewording of debugging text in manualroute to say "list of
1650 hosts" instead of "hostlist".
1652 2. If verify=header_syntax was set, and a header line with an unqualified
1653 address (no domain) and a large number of spaces between the end of the
1654 name and the colon was received, the reception process suffered a buffer
1655 overflow, and (when I tested it) crashed. This was caused by some obsolete
1656 code that should have been removed. The fix is to remove it!
1658 3. When running in the test harness, delay a bit after writing a bounce
1659 message to get a bit more predictability in the log output.
1661 4. Added a call to search_tidyup() just before forking a reception process. In
1662 theory, someone could use a lookup in the expansion of smtp_accept_max_
1663 per_host which, without the tidyup, could leave open a database connection.
1665 5. Added the variables $recipient_data and $sender_data which get set from a
1666 lookup success in an ACL "recipients" or "senders" condition, or a router
1667 "senders" option, similar to $domain_data and $local_part_data.
1669 6. Moved the writing of debug_print from before to after the "senders" test
1672 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1673 problems for message scanning, either using a data ACL, or using
1674 local_scan() because the Received: header was not generated till after they
1675 were called (in order to set the time as the time of reception completion).
1676 I have revised the way this works. The header is now generated after the
1677 body is received, but before the ACL or local_scan() are called. After they
1678 are run, the timestamp in the header is updated.
1684 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1685 before starting a queue runner without re-exec. This happened only when
1686 deliver_drop_privilege was set or when the Exim user was set to root. The
1687 effect of the bug was that timeouts during subsequent deliveries caused
1688 crashes instead of being properly handled. The handler is now left at its
1689 default (and expected) setting.
1691 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1692 message, again when deliver_drop_privilege is set or Exim is run as root.
1693 The bug described in (1) was not present in this case, but the tidying up
1694 of the other signals was missing. I have made the two cases consistent.
1696 3. The ignore_target_hosts setting on a manualroute router was being ignored
1697 for hosts that were looked up using the /MX notation.
1699 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1702 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1703 operated on the sender address. After changing the $sender_address to <>
1704 for the sender address verify, Exim was re-instated it as the original
1705 (before rewriting) address, but remembering that it had rewritten it, so it
1706 wasn't rewriting it again. This bug also had the effect of breaking the
1707 sender address verification caching when the sender address was rewritten.
1709 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1710 This has been changed so that if the ip literal address matches
1711 ignore_target_hosts, the router declines.
1713 7. Added expansion conditions match_domain, match_address, and match_local_
1714 part (NOT match_host).
1716 8. The placeholder for the Received: header didn't have a length field set.
1718 9. Added code to Exim itself and to exim_lock to test for a specific race
1719 condition that could lead to file corruption when using MBX delivery. The
1720 issue is with the lockfile that is created in /tmp. If this file is removed
1721 after a process has opened it but before that process has acquired a lock,
1722 there is the potential for a second process to recreate the file and also
1723 acquire a lock. This could lead to two Exim processes writing to the file
1724 at the same time. The added code performs the same test as UW imapd; it
1725 checks after acquiring the lock that its file descriptor still refers to
1726 the same named file.
1728 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1729 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1732 11. Added the smtp_active_hostname option. If used, this will typically be made
1733 to depend on the incoming interface address. Because $interface_address is
1734 not set up until the daemon has forked a reception process, error responses
1735 that can happen earlier (such as "too many connections") no longer contain
1738 12. If an expansion in a condition on a "warn" statement fails because a lookup
1739 defers, the "warn" statement is abandoned, and the next ACL statement is
1740 processed. Previously this caused the whole ACL to be aborted.
1742 13. Added the iplsearch lookup type.
1744 14. Added ident_timeout as a log selector.
1746 15. Added tls_certificate_verified as a log selector.
1748 16. Added a global option tls_require_ciphers (compare the smtp transport
1749 option of the same name). This controls incoming TLS connections.
1751 17. I finally figured out how to make tls_require_ciphers do a similar thing
1752 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1753 before starting the TLS session.
1755 18. Tabs are now shown as \t in -bP output.
1757 19. If the log selector return_path_on_delivery was set, Exim crashed when
1758 bouncing a message because it had too many Received: header lines.
1760 20. If two routers both had headers_remove settings, and the first one included
1761 a superfluous trailing colon, the final name in the first list and the
1762 first name in the second list were incorrectly joined into one item (with a
1763 colon in the middle).
1769 1. Added -C and -D options to the exinext utility, mainly to make it easier
1770 to include in the automated testing, but these could be helpful when
1771 multiple configurations are in use.
1773 2. The exinext utility was not formatting the output nicely when there was
1774 an alternate port involved in the retry record key, nor when there was a
1775 message id as well (for retries that were specific to a specific message
1776 and a specific host). It was also confused by IPv6 addresses, because of
1777 the additional colons they contain. I have fixed the IPv4 problem, and
1778 patched it up to do a reasonable job for IPv6.
1780 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1781 delivery, the log line now contains "pipelined" if PIPELINING was used.
1783 4. An SMTP transport process used to panic and die if the bind() call to set
1784 an explicit outgoing interface failed. This has been changed; it is now
1785 treated in the same way as a connect() failure.
1787 5. A reference to $sender_host_name in the part of a conditional expansion
1788 that was being skipped was still causing a DNS lookup. This no longer
1791 6. The def: expansion condition was not recognizing references to header lines
1792 that used bh_ and bheader_.
1794 7. Added the _cache feature to named lists.
1796 8. The code for checking quota_filecount in the appendfile transport was
1797 allowing one more file than it should have been.
1799 9. For compatibility with Sendmail, the command line option
1807 and sets the incoming protocol and host name (for trusted callers). The
1808 host name and its colon can be omitted when only the protocol is to be set.
1809 Note the Exim already has two private options, -pd and -ps, that refer to
1810 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1811 "s", but I don't think that's a major issue.
1813 10. A number of refactoring changes to the code, none of which should affect
1816 (a) The number of logging options was getting close to filling up the
1817 32-bit word that was used as a bit map. I have split them into two classes:
1818 those that are passed in the argument to log_write(), and those that are
1819 only ever tested independently outside of that function. These are now in
1820 separate 32-bit words, so there is plenty of room for expansion again.
1821 There is no change in the user interface or the logging behaviour.
1823 (b) When building, for example, log lines, the code previously used a
1824 macro that called string_cat() twice, in order to add two strings. This is
1825 not really sufficiently general. Furthermore, there was one instance where
1826 it was actually wrong because one of the argument was used twice, and in
1827 one call a function was used. (As it happened, calling the function twice
1828 did not affect the overall behaviour.) The macro has been replaced by a
1829 function that can join an arbitrary number of extra strings onto a growing
1832 (c) The code for expansion conditions now uses a table and a binary chop
1833 instead of a serial search (which was left over from when there were very
1834 few conditions). Also, it now recognizes conditions like "pam" even when
1835 the relevant support is not compiled in: a suitably worded error message is
1836 given if an attempt is made to use such a condition.
1838 11. Added ${time_interval:xxxxx}.
1840 12. A bug was causing one of the ddress fields not to be passed back correctly
1841 from remote delivery subprocesses. The field in question was not being
1842 subsequently used, so this caused to problems in practice.
1844 13. Added new log selectors queue_time and deliver_time.
1846 14. Might have fixed a bug in maildirsizefile handling that threw up
1847 "unexpected character" debug warnings, and recalculated the data
1848 unnecessarily. In any case, I expanded the warning message to give more
1851 15. Added the message "Restricted characters in address" to the statements in
1852 the default ACL that block characters like @ and % in local parts.
1854 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1855 Three changes have been made:
1857 (a) There was a serious bug; a negative response to MAIL caused the whole
1858 recipient domain to be cached as invalid, thereby blocking all messages
1859 to all local parts at the same domain, from all senders. This bug has
1860 been fixed. The domain is no longer cached after a negative response to
1861 MAIL if the sender used is not empty.
1863 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1866 (c) A new callout option, "use_sender" has been added for people who want
1867 the modified behaviour.
1873 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1876 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1877 indeed breaks things for older releases.
1879 3. Added additional logging to the case where there is a problem reading data
1880 from a filter that is running in a subprocess using a pipe, in order to
1881 try to track down a specific problem.
1883 4. Testing facility fudge: when running in the test harness and attempting
1884 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1885 getting "No route to host". Convert this to a timeout.
1887 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1890 6. Some OS don't have socklen_t but use size_t instead. This affects the
1891 fifth argument of getsockopt() amongst other things. This is now
1892 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1893 can be set for individual OS. I have set it for SunOS5, OSF1, and
1894 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1895 some earlier ones do not.
1897 7. Change 4.30/15 was not doing the test caselessly.
1899 8. The standard form for an IPv6 address literal was being rejected by address
1900 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1901 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1902 this, as well as the form without the "IPv6" on the front (but only when
1903 address literals are enabled, of course).
1905 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1907 10. Exim crashed if a message with an empty sender address specified by -f
1908 encountered a router with an errors_to setting. This could be provoked only
1909 by a command such as
1913 where an empty string was supplied; "<>" did not hit this bug.
1915 11. Installed PCRE release 4.5.
1917 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1918 remained set. It is now erased.
1920 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1921 times from message ids (which are base 36 rather than the normal 62).
1923 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1924 were being counted as actual protocol errors, and logged if the log
1925 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1926 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1927 and DATA following a set of rejected RCPTs do not count as protocol errors.
1928 In other words, Exim assumes they were pipelined, though this may not
1929 actually be the case. Of course, in all cases the client gets an
1930 appropriate error code.
1932 15. If a lookup fails in an ACL condition, a message about the failure may
1933 be available; it is used if testing the ACL cannot continue, because most
1934 such messages specify what the cause of the deferral is. However, some
1935 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1936 that caused an old message to be retained and used if a later statement
1937 caused a defer, replacing the real cause of the deferral.
1939 16. If an IP address had so many PTR records that the DNS lookup buffer
1940 was not large enough to hold them, Exim could crash while trying to process
1941 the truncated data. It now detects and logs this case.
1943 17. Further to 4.21/58, another change has been made: if (and only if) the
1944 first line of a message (the first header line) ends with CRLF, a bare LF
1945 in a subsequent header line has a space inserted after it, so as not to
1946 terminate the header.
1948 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1949 unnecessarily, used atoi() instead of strtol(), and didn't check the
1950 termination when getting file sizes from file names by regex.
1952 19. Completely re-implemented the support for maildirsize files, in the light
1953 of a number of problems with the previous contributed implementation
1954 (4.30/29). In particular:
1956 . If the quota is zero, the maildirsize file is maintained, but no quota is
1959 . If the maildir directory does not exist, it is created before any attempt
1960 to write a maildirsize file.
1962 . The quota value in the file is just a cache; if the quota is changed in
1963 the transport, the new value overrides.
1965 . A regular expression is available for excluding directories from the
1968 20. The autoreply transport checks the characters in options that define the
1969 message's headers; it allows continued headers, but it was checking with
1970 isspace() after an embedded newline instead of explicitly looking for a
1973 21. If all the "regular" hosts to which an address was routed had passed their
1974 expiry times, and had not reached their retry times, the address was
1975 bounced, even if fallback hosts were defined. Now Exim should go on to try
1978 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1979 equivalent code in the SMTP transport. Some hosts send humungous responses
1980 to HELO/EHLO, more than 1024 it seems.
1982 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1983 gives compiler warnings in some environments. I've now done it "properly",
1986 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1987 (because of problems with the built-in one) was declared to return uschar *
1988 instead of char *, causing compiler failure.
1990 25. Fixed a file descriptor leak when processing alias/forward files.
1992 26. Fixed a minor format string issue in dbfn.c.
1994 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1996 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1997 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2000 29. When a sender address is verified, it is cached, to save repeating the test
2001 when there is more than one recipient in a message. However, when the
2002 verification involves a callout, it is possible for different callout
2003 options to be set for different recipients. It is too complicated to keep
2004 track of this in the cache, so now Exim always runs a verification when a
2005 callout is required, relying on the callout cache for the optimization.
2006 The overhead is duplication of the address routing, but this should not be
2009 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2010 to be verified with callout=postmaster, and the main callout worked but the
2011 postmaster check failed, the verification correctly failed. However, if a
2012 subsequent RCPT command asked for sender verification *without* the
2013 postmaster check, incorrect caching caused this verification also to fail,
2016 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2017 it was not caching the DNS options (qualify_single, search_parents) that
2018 were used when the lookup failed. A subsequent lookup with different
2019 options therefore always gave the same answer, though there were cases
2020 where it should not have. (Example: a "domains = !$mx_any" option on a
2021 dnslookup router: the "domains" option is always processed without any
2022 widening, but the router might have qualify_single set.) Now Exim uses the
2023 cached value only when the same options are set.
2025 32. Added John Jetmore's "exipick" utility to the distribution.
2027 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2028 than a timeout (e.g. a certificate is required, and is not provided), an
2029 Exim server now closes the connection immediately. Previously it waited for
2030 the client to close - but if the client is SSL, it seems that they each
2031 wait for each other, leading to a delay before one of them times out.
2033 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2034 maintained 0.8.x compatibility because I don't think many are using it, and
2035 it is clearly obsolete.
2037 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2040 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2041 client certificate was expired. A simple patch fixes this, though I don't
2042 understand the full logic of why the verify callback is called multiple
2045 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2046 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2047 which causes problems with some clients (such as the Certicom SSL Plus
2048 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2049 disables the coutermeasure allowing Eudora to connect."
2051 38. Exim was not checking that a write() to a log file succeeded. This could
2052 lead to Bad Things if a log got too big, in particular if it hit a file
2053 size limit. Exim now panics and dies if it cannot write to a log file, just
2054 as it does if it cannot open a log file.
2056 39. Modified OS/Makefile-Linux so that it now contains
2058 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2060 The two -D definitions ensure that Exim is compiled with large file
2061 support, which makes it possible to handle log files that are bigger than
2064 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2065 instance) a domain was checked against a named list that involved a lookup,
2066 causing $domain_data to be set, then another domain was checked against the
2067 same list, then the first domain was re-checked, the value of $domain_data
2068 after the final check could be wrong. In particular, if the second check
2069 failed, it could be set empty. This bug probably also applied to
2072 41. The strip_trailing_dot option was not being applied to the address given
2073 with the -f command-line option.
2075 42. The code for reading a message's header from the spool was incrementing
2076 $received_count, but never initializing it. This meant that the value was
2077 incorrect (doubled) while delivering a message in the same process in which
2078 it was received. In the most common configuration of Exim, this never
2079 happens - a fresh exec is done - but it can happen when
2080 deliver_drop_privilege is set.
2082 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2083 it now includes up to 150 characters of the unexpected data in the log
2086 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2087 and building data strings. The size of both of these buffers was 10 000
2088 bytes - far larger than anybody would *ever* want, thought I. Needless to
2089 say, somebody hit the limit. I have increased the maximum line length to
2090 20 000 and the maximum data length of concatenated lines to 100 000. I have
2091 also fixed two bugs, because there was no checking on these buffers. Tsk,
2092 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2093 buffer is too small.
2095 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2096 lsearch lookups. Now it does.
2098 46. When parsing a route_list item in a manualroute router, a fixed-length
2099 buffer was used for the list of hosts. I made this 1024 bytes long,
2100 thinking that nobody would ever have a list of hosts that long. Wrong.
2101 Somebody had a whole pile of complicated expansion conditions, and the
2102 string was silently truncated, leading to an expansion error. It turns out
2103 that it is easier to change to an unlimited length (owing to other changes
2104 that have happened since this code was originally written) than to build
2105 structure for giving a limitation error. The length of the item that
2106 expands into the list of hosts is now unlimited.
2108 47. The lsearch lookup could not handle data where the length of text line was
2109 more than 4095 characters. Such lines were truncated, leading to shortened
2110 data being returned. It should now handle lines of any length.
2112 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2113 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2116 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2117 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2118 no longer happens in comment lines. A list of replacements is now placed
2119 at the head of all of the source files, except those whose only change is
2120 to replace PERL_COMMAND in the very first #! line.
2122 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2123 messages on the queue, with a bottom-up merge sort, using code contributed
2124 by Michael Haardt. This should make operations like -bp somewhat faster on
2125 large queues. It won't affect queue runners, except when queue_run_in_order
2128 51. Installed eximstats 1.31 in the distribution.
2130 52. Added support for SRV lookups to the dnslookup router.
2132 53. If an ACL referred to $message_body or $message_body_end, the value was not
2133 reset for any messages that followed in the same SMTP session.
2135 54. The store-handling optimization for building very long strings was not
2136 differentiating between the different store pools. I don't think this
2137 actually made any difference in practice, but I've tidied it.
2139 55. While running the routers to verify a sender address, $sender_address
2140 was still set to the sender address. This is wrong, because when routing to
2141 send a bounce to the sender, it would be empty. Therefore, I have changed
2142 it so that, while verifying a sender address, $sender_address is set to <>.
2143 (There is no change to what happens when verifying a recipient address.)
2145 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2146 target A or AAAA records (if not already returned) without resetting the
2147 qualify_single or search_parents options of the DNS resolver. These are
2148 inappropriate in this case because the targets of MX and SRV records must
2149 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2150 target that, when qualified, matched something in the local domain. These
2151 two options are now turned off when doing these lookups.
2153 57. It seems that at least some releases of Reiserfs (which does not have the
2154 concept of a fixed number of inodes) returns zero and not -1 for the
2155 number of available inodes. This interacted badly with check_spool_inodes,
2156 which assumed that -1 was the "no such thing" setting. What I have done is
2157 to check that the total number of inodes is greater than zero before doing
2158 the test of how many are available.
2160 58. When a "warn" ACL statement has a log_message modifier, the message is
2161 remembered, and not repeated. This is to avoid a lot of repetition when a
2162 message has many recipients that cause the same warning to be written.
2163 Howewer, Exim was preserving the list of already written lines for an
2164 entire SMTP session, which doesn't seem right. The memory is now reset if a
2165 new message is started.
2167 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2168 debugging output unless log_rewrite was also set.
2170 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2171 of (int)(handle) when we know that handle contains (void *)(-1).
2173 61. The Exim daemon panic-logs an error return when it closes the incoming
2174 connection. However "connection reset by peer" seems to be common, and
2175 isn't really an error worthy of noting specially, so that particular error
2178 62. When Exim is trying to find all the local interfaces, it used to panic and
2179 die if the ioctl to get the interface flags failed. However, it seems that
2180 on at least one OS (Solaris 9) it is possible to have an interface that is
2181 included in the list of interfaces, but for which you get a failure error
2182 for this call. This happens when the interface is not "plumbed" into a
2183 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2184 failure of the "get flags" call assumes that the interface is down.
2186 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2187 makes life easier for people who are doing arithmetic on fields extracted
2188 from dates, where you often get leading zeros that should not be
2189 interpreted as octal.
2191 64. Added qualify_domain to the redirect router, to override the global
2194 65. If a pathologically long header line contained very many addresses (the
2195 report of this problem mentioned 10 000) and each of them was rewritten,
2196 Exim could use up a very large amount of memory. (It kept on making new
2197 copies of the header line as it rewrote, and never released the old ones.)
2198 At the expense of a bit more processing, the header rewriting function has
2199 been changed so that it no longer eats memory in this way.
2201 66. The generation of the Received: header has been moved from the time that a
2202 message starts to be received, to the time that it finishes. The timestamp
2203 in the Received: header should now be very close to that of the <= log
2204 line. There are two side-effects of this change:
2206 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2207 logged header lines no longer include the local Received: line, because
2208 it has not yet been created. The same applies to a copy of the message
2209 that is returned to a non-SMTP sender when a message is rejected.
2211 (b) When a filter file is tested using -bf, no additional Received: header
2212 is added to the test message. After some thought, I decided that this
2215 This change does not affect the value of $received_for. It is still set
2216 after address rewriting, but before local_scan() is called.
2218 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2220 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2221 gave an unhelpful panic error message, and a defer error. I have managed to
2222 change this behaviour so that it now rejects any supplied certificate,
2223 which seems right, as the list of acceptable certificates is empty.
2225 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2226 gave an unhelpful defer error. I have not managed to make this reject any
2227 supplied certificates, but the error message it gives is "no certificate
2228 supplied", which is not helpful.
2230 70. exigrep's output now also includes lines that are not associated with any
2231 message, but which match the given pattern. Implemented by a patch from
2232 Martin Sluka, which also tidied up the Perl a bit.
2234 71. Recipient callout verification, like sender verification, was using <> in
2235 the MAIL FROM command. This isn't really the right thing, since the actual
2236 sender may affect whether the remote host accepts the recipient or not. I
2237 have changed it to use the actual sender in the callout; this means that
2238 the cache record is now keyed on a recipient/sender pair, not just the
2239 recipient address. There doesn't seem to be a real danger of callout loops,
2240 since a callout by the remote host to check the sender would use <>.
2241 [SEE ABOVE: changed after hitting problems.]
2243 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2244 temporary errors. However, in the case of such a code being given after
2245 the end of a data transmission (i.e. after ".") Exim was failing to write
2246 a retry record for the message. (Yes, there was some broken host that was
2247 actually sending 8xx at this point.)
2249 73. An unknown lookup type in a host list could cause Exim to panic-die when
2250 the list was checked. (An example that provoked this was putting <; in the
2251 middle of a list instead of at the start.) If this happened during a DATA
2252 ACL check, a -D file could be left lying around. This kind of configuration
2253 error no longer causes Exim to die; instead it causes a defer errror. The
2254 incident is still logged to the main and panic logs.
2256 74. Buglet left over from Exim 3 conversion. The message "too many messages
2257 in one connection" was written to the rejectlog but not the mainlog, except
2258 when address rewriting (yes!) was being logged.
2260 75. Added write_rejectlog option.
2262 76. When a system filter was run not as root (that is, when system_filter_user
2263 was set), the values of the $n variables were not being returned to the
2264 main process; thus, they were not subsequently available in the $sn
2267 77. Added +return_path_on_delivery log selector.
2269 78. A connection timeout was being treated differently from recipients deferred
2270 when testing hosts_max_try with a message that was older than the host's
2271 retry timeout. (The host should not be counted, thus allowing all hosts to
2272 be tried at least once before bouncing.) This may have been the cause of an
2273 occasionally reported bug whereby a message would remain on the queue
2274 longer than the retry timeout, but would be bounced if a delivery was
2275 forced. I say "may" because I never totally pinned down the problem;
2276 setting up timeout/retry tests is difficult. See also the next item.
2278 79. The ultimate address timeout was not being applied to errors that involved
2279 a combination of host plus message (for example, a timeout on a MAIL
2280 command). When an address resolved to a number of possible hosts, and they
2281 were not all tried for each delivery (e.g. because of hosts_max_try), a
2282 message could remain on the queue longer than the retry timeout.
2284 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2287 81. Fixed an obscure SMTP outgoing bug which required at least the following
2288 conditions: (a) there was another message waiting for the same server;
2289 (b) the server returned 5xx to all RCPT commands in the first message so
2290 that the message was not completed; (c) the server dropped the connection
2291 or gave a negative response to the RSET that Exim sends to abort the
2292 transaction. The observed case was a dropped connection after DATA that had
2293 been sent in pipelining mode. That is, the server had advertised PIPELINING
2294 but was not implementing it correctly. The effect of the bug was incorrect
2295 behaviour, such as trying another host, and this could lead to a crash.
2301 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2302 and daemon.c were passed as pointers to ints; they should have been
2303 pointers to socklen_t variables (which are typically unsigned ints).
2305 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2308 3. Fixed a really odd bug that affected only the testing scheme; patching a
2309 certain fixed string in the binary changed the value of another string that
2310 happened to be identical to the end of the original first string.
2312 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2313 name", it returns that address as the IP address. On some operating
2314 systems (e.g. Solaris), it also passes back the IP address string as the
2315 "host name". However, on others (e.g. Linux), it passes back an empty
2316 string. Exim wasn't checking for this, and was changing the host name to an
2317 empty string, assuming it had been canonicized.
2319 5. Although rare, it is permitted to have more than one PTR record for a given
2320 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2321 all the names associated with an address, because they do in Solaris.
2322 However, it seems that they do not in Linux for data that comes from the
2323 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2324 I found this out when I moved to a new Linux workstation and tried to run
2325 the Exim test suite.
2327 To get round this problem I have changed the code so that it now does its
2328 own call to the DNS to look up PTR records when searching for a host name.
2329 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2330 addresses that are only in /etc/hosts are still found.
2332 This behaviour is, however, controlled by an option called host_lookup_
2333 order, which defaults to "bydns:byaddr". If people want to use the other
2334 order, or indeed, just use one or the other means of lookup, they can
2335 specify it in this variable.
2337 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2338 some operating systems, this comes back from gethostbyaddr() as an empty
2339 string, and this is what Exim used to test for. However, it seems that in
2340 other systems, "." is yielded. Exim now tests for this case too.
2342 7. The values of check_spool_space and check_log_space are now held internally
2343 as a number of kilobytes instead of an absolute number of bytes. If a
2344 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2345 kilobyte. This means that much larger values can be stored.
2347 8. Exim monitor: an attempt to get the action menu when not actually pointing
2348 at a message produces an empty menu entitled "No message selected". This
2349 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2350 no entries in it ("Shell widget menu has zero width and/or height"). So I
2351 have added a single, blank menu entry in this case.
2353 9. Added ${quote_local_part.
2355 10. MIME decoding is now applied to the contents of Subject: header lines when
2358 11. Now that a reference to $sender_host_address automatically causes a reverse
2359 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2360 host lookup before query-style lookups in lists that might use this
2361 variable. This has therefore been abolished, and the "net-" prefix is no
2362 longer necessary for query-style lookups.
2364 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2365 have been a typo for LFLAGS, so it has been changed.
2367 13. The install script calls Exim with "-C /dev/null" in order to find the
2368 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2369 to be output. Howeve, since Exim outputs its version number before the
2370 error, it didn't break the script. It just looked ugly. I fixed this by
2371 always allowing "-C /dev/null" if the caller is root.
2373 14. Ignore overlarge ACL variable number when reading spool file - insurance
2374 against a later release with more variables having written the file.
2376 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2377 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2378 this, as well as the form without the "IPv6" on the front.
2380 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2381 OS/Makefile-Darwin file.
2383 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2384 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2386 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2387 variable. However, some OS use ints for inodes. I've added cast to long int
2388 to get rid of the compiler warning.
2390 19. I had forgotten to lock out "/../" in configuration file names when
2391 ALT_CONFIG_PREFIX was set.
2393 20. Routers used for verification do not need to specify transports. However,
2394 if such a router generated a host list, and callout was configured, Exim
2395 crashed, because it could not find a port number from the (non-existent)
2396 transport. It now assumes port 25 in this circumstance.
2398 21. Added the -t option to exigrep.
2400 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2401 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2402 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2403 them is set, LOOKUP_LSEARCH is forced.
2405 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2407 24. Added sender and host information to the "rejected by local_scan()" log
2408 line; previously there was no indication of these.
2410 25. Added .include_if_exists.
2412 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2413 Linux. It turns out that not all file systems support this. Apparently some
2414 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2415 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2416 not supported on the file descriptor, is now ignored when Exim is trying to
2417 sync a directory. This applies only to Linux.
2419 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2421 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2422 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2425 29. Added support for maildirsize files from supplied patch (modified a bit).
2427 30. The use of :fail: followed by an empty string could lead Exim to respond to
2428 sender verification failures with (e.g.):
2430 550 Verification failed for <xxx>
2431 550 Sender verify failed
2433 where the first response line was missing the '-' that indicates it is not
2434 the final line of the response.
2436 31. The loop for finding the name of the user that called Exim had a hardwired
2437 limit of 10; it now uses the value of finduser_retries, which is used for
2438 all other user lookups.
2440 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2443 33. Exim was neglecting to zero errno before one call of strtol() when
2444 expanding a string and expecting an integer value. On some systems this
2445 resulted in spurious "integer overflow" errors. Also, it was casting the
2446 result into an int without checking.
2448 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2449 did not work. The code looks as if it has *never* worked, though it appears
2450 to have been documented since at least releast 1.62. I have made it work.
2452 35. The "timeout_DNS" error in retry rules, also documented since at least
2453 1.62, also never worked. As it isn't clear exactly what this means, and
2454 clearly it isn't a major issue, I have abolished the feature by treating it
2455 as "timeout", and writing a warning to the main and panic logs.
2457 36. The display of retry rules for -brt wasn't always showing the error code
2460 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2461 timeout_connect_A, timeout_connect_MX.
2463 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2464 to the empty sender.
2466 39. The daemon was not analysing the content of -oX till after it had closed
2467 stderr and disconnected from the controlling terminal. This meant that any
2468 syntax errors were only noted on the panic log, and the return code from
2469 the command was 0. By re-arranging the code a little, I've made the
2470 decoding happen first, so such errors now appear on stderr, and the return
2471 code is 1. However, the actual setting up of the sockets still happens in
2472 the disconnected process, so errors there are still only recorded on the
2475 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2476 connections (as happens on some IP stacks) was logged at start up time as
2477 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2478 it from "IPv6 and IPv4", which means that two separate sockets are being
2481 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2482 says whether AF_INET or AF_INET6 was passed as an argument.
2484 42. Exiwhat output was messed up when time zones were included in log
2487 43. Exiwhat now gives more information about the daemon's listening ports,
2488 and whether -tls-on-connect was used.
2490 44. The "port" option of the smtp transport is now expanded.
2492 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2493 silently ignored. Now an error message is written to the main and panic
2496 46. There's a new ACL modifier called "logwrite" which writes to a log file
2497 as soon as it is encountered.
2499 47. Added $local_user_uid and $local_user_gid at routing time.
2501 48. Exim crashed when trying to verify a sender address that was being
2504 49. Exim was recognizing only a space character after ".include". It now also
2505 recognizes a tab character.
2507 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2508 extracting the relevant information from the specification. The man page no
2509 longer contains scrambled data for the -d option, and I've added a section
2510 at the front about calling Exim under different names.
2512 51. Added "extra_headers" argument to the "mail" command in filter files.
2514 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2517 53. Installed eximstats 1.29.
2519 54. Added transport_filter_timeout as a generic transport option.
2521 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2522 Cc: header lines. This was required by RFC 822, but it not required by RFC
2525 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2526 incoming messages that did not have them. Now it does so only if the
2527 message originates locally, that is, if there is no associated remote host
2528 address. When Resent- header lines are present, this applies to the Resent-
2529 lines rather than the non-Resent- lines.
2531 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2532 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2534 58. Messages for configuration errors now include the name of the main
2535 configuration file - useful now that there may be more than one file in a
2536 list (.included file names were always shown).
2538 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2539 for those rare installations that do not start the daemon as root or run it
2540 setuid root. I've cut out the call to initgroups() if the daemon is not
2543 60. The Exim user and group can now be bound into the binary as text strings
2544 that are looked up at the start of Exim's processing.
2546 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2548 62. Added $mailstore_basename variable.
2550 63. Installed patch to sieve.c from Michael Haardt.
2552 64. When Exim failed to open the panic log after failing to open the main log,
2553 the original message it was trying to log was written to stderr and debug
2554 output, but if they were not available (the usual case in production), it
2555 was lost. Now it is written to syslog before the two lines that record the
2556 failures to open the logs.
2558 65. Users' Exim filters run in subprocesses under the user's uid. It is
2559 possible for a "deliver" command or an alias in a "personal" command to
2560 provoke an address rewrite. If logging of address rewriting is configured,
2561 this fails because the process is not running as root or exim. There may be
2562 a better way of dealing with this, but for the moment (because 4.30 needs
2563 to be released), I have disabled address rewrite logging when running a
2564 filter in a non-root, non-exim process.
2570 1. The buildconfig auxiliary program wasn't quoting the value set for
2571 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2572 not defined. This bug was masked in 4.22 by the effect that was fixed in
2575 2. Some messages that were rejected after a message id was allocated were
2576 shown as "incomplete" by exigrep. It no longer does this for messages that
2577 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2579 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2580 have allow_domain_literals set, the ID did not get logged in the <= line.
2581 Domain literals are now always recognized in Message-ID: header lines.
2583 4. The first argument for a ${extract expansion item is the key name or field
2584 number. Leading and trailing spaces in this item were not being ignored,
2585 causing some misleading effects.
2587 5. When deliver_drop_privilege was set, single queue runner processes started
2588 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2589 same command in the process it spins off) were not dropping privilege.
2591 6. When the daemon running as "exim" started a queue runner, it always
2592 re-executed Exim in the spun-off process. This is a waste of effort when
2593 deliver_drop_privilege is set. The new process now just calls the
2594 queue-runner function directly.
2600 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2603 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2604 ignored. Though the use of -oP was forcing the writing of a pid file, it
2605 was always written to the default place.
2607 3. If the message "no IP address found for host xxxx" is generated during
2608 incoming verification, it is now followed by identification of the incoming
2609 connection (so you can more easily find what provoked it).
2611 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2613 5. Added some features to "harden" Exim a bit more against certain attacks:
2615 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2616 be put in Local/Makefile. This is like the never_users runtime option,
2617 but it cannot be overridden. The default setting is "root".
2619 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2620 prefix string with which any file named in a -C command line option
2623 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2624 is retained for -C and -D only if the caller of Exim is root. Without
2625 it, the exim user may also use -C and -D and retain privilege.
2627 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2628 command line option is disabled.
2630 6. Macro names set by the -D option must start with an upper case letter, just
2631 like macro names defined in the configuration file.
2633 7. Added "dereference=" facility to LDAP.
2635 8. Two instances of the typo "uknown" in the source files are fixed.
2637 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2638 the Configure-Makefile script screwed up while processing it.
2640 10. Incorporated PCRE 4.4.
2642 11. The SMTP synchronization check was not operating right at the start of an
2643 SMTP session. For example, it could not catch a HELO sent before the client
2644 waited for the greeting. There is now a check for outstanding input at the
2645 point when the greeting is written. Because of the duplex, asynchronous
2646 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2647 way, but not yet received, when the check is performed.
2649 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2650 on TCP/IP sockets, because this apparently causes some broken clients to
2653 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2654 unchanged) from the Cygwin maintainer.
2656 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2657 was supported instead of testing for mbx. Effectively a typo.
2659 15. The spa authenticator server code was not checking that the input it
2660 received was valid base64.
2662 16. The debug output line for the "set" modifier in ACLs was not showing the
2663 name of the variable that was being set.
2665 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2667 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2668 lookup on its own. Something else was needed to trigger the lookup. For
2669 example, a match in host_lookup or the need for a host name in a host list.
2670 Now, if $sender_host_name is referenced and the host name has not yet been
2671 looked up, a lookup is performed. If the lookup fails, the variable remains
2672 empty, and $host_lookup_failed is set to "1".
2674 19. Added "eqi" as a case-independent comparison operator.
2676 20. The saslauthd authentication condition could segfault if neither service
2677 nor realm was specified.
2679 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2680 error message that was logged was misleading, and incoming SMTP
2681 connections were dropped. The message is now more accurate, and temporary
2682 errors are given to SMTP connections.
2684 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2685 (except RSET) with a 421 error, until QUIT is received. However, it was
2686 failing to send a response to QUIT.
2688 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2689 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2690 ACL is now run after the helo_try_verify_hosts code.
2692 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2693 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2694 case-independent, so other case variants are also recognized). Apparently
2695 some systems use these upper case variants.
2697 25. If more than two messages were waiting for the same host, and a transport
2698 filter was specified for the transport, Exim sent two messages over the
2699 same TCP/IP connection, and then failed with "socket operation on non-
2700 socket" when it tried to send the third.
2702 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2704 27. The extern definition of crypt16() in expand.c was not being excluded when
2705 the OS had its own crypt16() function.
2707 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2708 as a preferred synonym for return_size_limit, both as an option and as an
2711 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2713 30. Changed the default configuration ACL to relax the local part checking rule
2714 for addresses that are not in any local domains. For these addresses,
2715 slashes and pipe symbols are allowed within local parts, but the sequence
2716 /../ is explicitly forbidden.
2718 31. SPA server authentication was not clearing the challenge buffer before
2721 32. log_message in a "warn" ACL statement was writing to the reject log as
2722 well as to the main log, which contradicts the documentation and doesn't
2723 seem right (because no rejection is happening). So I have stopped it.
2725 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2726 However, I am unable to do any testing of this.
2728 34. Fixed an infelicity in the appendfile transport. When checking directories
2729 for a mailbox, to see if any needed to be created, it was accidentally
2730 using path names with one or more superfluous leading slashes; tracing
2731 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2733 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2734 contents are added to the log line that is written for every discarded
2735 recipient. (Previously a log_message setting was ignored.)
2737 36. The ${quote: operator now quotes the string if it is empty.
2739 37. The install script runs exim in order to find its version number. If for
2740 some reason other than non-existence or emptiness, which it checks, it
2741 could not run './exim', it was installing it with an empty version number,
2742 i.e. as "exim-". This error state is now caught, and the installation is
2745 38. An argument was missing from the function that creates an error message
2746 when Exim fails to connect to the socket for saslauthd authentication.
2747 This could cause Exim to crash, or give a corrupted message.
2749 39. Added isip, isip4, and isip6 to ${if conditions.
2751 40. The ACL variables $acl_xx are now saved with the message, and can be
2752 accessed later in routers, transports, and filters.
2754 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2755 strings in the file are not string-expanded.
2757 42. If a MAIL command specified a SIZE value that was too large to fit into an
2758 int variable, the check against message_size_limit failed. Such values are
2759 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2760 day this will have to be increased, but I don't think I want to be around
2761 when emails are that large.
2768 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2769 iconv() is not standard in FreeBSD.
2771 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2772 IPv6 enabled. The observed symptom was a segmentation fault on return from
2773 the function os_common_find_running_interfaces() in src/os.c.
2775 3. In the check_special_case() function in daemon.c I had used "errno" as an
2776 argument name, which causes warnings on some systems. This was basically a
2777 typo, since it was named "eno" in the comments!
2779 4. The code that waits for the clock to tick (at a resolution of some fraction
2780 of a second) so as to ensure message-id uniqueness was always waiting for
2781 at least one whole tick, when it could have waited for less. [This is
2782 almost certainly not relevant at current processor speeds, where it is
2783 unlikely to ever wait at all. But we try to future-proof.]
2785 5. The function that sleeps for a time interval that includes fractions of a
2786 second contained a race. It did not block SIGALRM between setting the
2787 timer, and suspending (a couple of lines later). If the interval was short
2788 and the sigsuspend() was delayed until after it had expired, the suspension
2789 never ended. On busy systems this could lead to processes getting stuck for
2792 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2793 process, before it forks any delivery processes. The open lookup caching
2794 mechanism meant that the open file or database connection was passed into
2795 the delivery process. The problem was that delivery processes always tidy
2796 up cached lookup data. This could cause a problem for the next delivery
2797 process started by the queue runner, because the external queue runner
2798 process does not know about the closure. So the next delivery process
2799 still has data in the lookup cache. In the case of a file lookup, there was
2800 no problem because closing a file descriptor in a subprocess doesn't affect
2801 the parent. However, if the lookup was caching a connection to a database,
2802 the connection was closed, and the second delivery process was likely to
2803 see errors such as "PGSQL: query failed: server closed the connection
2804 unexpectedly". The problem has been fixed by closing all cached lookups
2805 in a queue runner before running a delivery process.
2807 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2808 seem to have the "const" qualifier which it has on other OS. I've
2811 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2812 *of the same type* (client or server) with the same public name that an
2813 error should be diagnosed.
2815 9. When Exim looked up a host name for an IP address, but failed to find the
2816 original IP address when looking up the host name (a safety check), it
2817 output the message "<ip address> does not match any IP for NULL", which was
2818 confusing, to say the least. The bug was that the host name should have
2819 appeared instead of "NULL".
2821 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2822 user that is built into the binary, and the -C or -D options is used, root
2823 privilege is dropped before the configuration file is read. In addition,
2824 logging is switched to stderr instead of the normal log files. If the
2825 configuration then re-defines the Exim user, the unprivileged environment
2826 is probably not what is expected, so Exim logs a panic warning message (but
2829 However, if deliver_drop_privilege is set, the unprivileged state may well
2830 be exactly what is intended, so the warning has been cut out in that case,
2831 and Exim is allowed to try to write to its normal log files.
2837 1. smtp_return_error_details was not giving details for temporary sender
2838 or receiver verification errors.
2840 2. Diagnose a configuration error if two authenticators have the same public
2843 3. Exim used not to create the message log file for a message until the first
2844 delivery attempt. This could be confusing when incoming messages were held
2845 for policy or load reasons. The message log file is now created at the time
2846 the message is received, and an initial "Received" line is written to it.
2848 4. The automatically generated man page for command line options had a minor
2849 bug that caused no ill effects; however, a more serious problem was that
2850 the procedure for building the man page automatically didn't always
2851 operate. Consequently, release 4.20 contains an out-of-date version. This
2852 shouldn't happen again.
2854 5. When building Exim with embedded Perl support, the script that builds the
2855 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2856 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2858 6. The freeze_tell option was not being used for messages that were frozen on
2859 arrival, either by an ACL or by local_scan().
2861 7. Added the smtp_incomplete_transaction log selector.
2863 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2864 was accepting AUTH without a new EHLO.
2866 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2867 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2870 10. Logging of TCP/IP connections (when configured) now happens in the main
2871 daemon process instead of the child process, so that the TCP/IP connection
2872 count is more accurate (but it can never be perfect).
2874 11. The use of "drop" in a nested ACL was not being handled correctly in the
2875 outer ACL. Now, if condition failure induced by the nested "drop" causes
2876 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2877 or "require"), the connection is dropped.
2879 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2880 that yield "discard" can now be used with an "accept" or a "discard" verb,
2881 but an error is generated for any others (because I can't see a useful way
2882 to define what should happen).
2884 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2885 are now processed in the same way as lines in the Exim configuration file.
2886 In particular, continuation lines are supported.
2888 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2890 15. Added -ti meaning -t -i.
2892 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2893 domains, and warn by logging if others are found.
2895 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2896 returned from GIFCONF when Exim is trying to find the list of interfaces on
2897 a host. The code in os.c has been modified to copy each ifreq to an aligned
2898 structure in all cases.
2900 Also, in some cases, the returned ifreq's were being copied to a 'struct
2901 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2902 means the last couple of bytes of an IPv6 address could be chopped if the
2903 ifreq contained only a normal sockaddr (14 bytes storage).
2905 18. Named domain lists were not supported in the hosts_treat_as_local option.
2906 An entry such as +xxxx was not recognized, and was treated as a literal
2909 19. Ensure that header lines added by a DATA ACL are included in the reject log
2910 if the ACL subsequently rejects the message.
2912 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2913 MD5 (which is deprecated).
2915 21. When testing a filter file using -bf, Exim was writing a message when it
2916 took the sender from a "From " line in the message, but it was not doing so
2917 when it took $return_path from a Return-Path: header line. It now does.
2919 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2920 with a valid header line field name (a series of printing characters
2921 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2923 23. Changed "disc" in the source to "disk" to conform to the documentation and
2924 the book and for uniformity.
2926 24. Ignore Sendmail's -Ooption=value command line item.
2928 25. When execve() failed while trying to run a command in a pipe transport,
2929 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2930 could be confused with a return value of 69 from the command itself. This
2931 has been changed to 127, the value the shell returns if it is asked to run
2932 a non-existent command. The wording for the related log line suggests a
2933 non-existent command as the problem.
2935 26. If received_header_text expands to an empty string, do not add a Received:
2936 header line to the message. (Well, it adds a token one on the spool, but
2937 marks it "old" so that it doesn't get used or transmitted.)
2939 27. Installed eximstats 1.28 (addition of -nt option).
2941 28. There was no check for failure on the call to getsockname() in the daemon
2942 code. This can fail if there is a shortage of resources on the system, with
2943 ENOMEM, for example. A temporary error is now given on failure.
2945 29. Contrary to the C standard, it seems that in some environments, the
2946 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2947 program. Exim now does this explicitly; it affects the formatting of
2948 timestamps using strftime().
2950 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2951 complaints. I've now initialized all the variables, to avoid this.
2953 32. Header lines added by a system filter were not being "seen" during
2954 transport-time rewrites.
2956 33. The info_callback() function passed to OpenSSL is set up with type void
2957 (*)(SSL *, int, int), as described somewhere. However, when calling the
2958 function (actually a macro) that sets it up, the type void(*)() is
2959 expected. I've put in a cast to prevent warnings from picky compilers.
2961 34. If a DNS black list lookup found a CNAME record, but there were no A
2962 records associated with the domain it pointed at, Exim crashed.
2964 35. If a DNS black list lookup returned more than one A record, Exim ignored
2965 all but the first. It now scans all returned addresses if a particular IP
2966 value is being sought. In this situation, the contents of the
2967 $dnslist_value variable are a list of all the addresses, separated by a
2970 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2971 to accept a host name and all its aliases if the forward lookup for any of
2972 them yielded the IP address of the incoming connection. Now it accepts only
2973 those names whose forward lookup yields the correct IP address. Any other
2974 names are discarded. This closes a loophole whereby a rogue DNS
2975 administrator could create reverse DNS records to break through a
2976 wildcarded host restriction in an ACL.
2978 37. If a user filter or a system filter that ran in a subprocess used any of
2979 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2980 the wrong values were passed to the pipe command ($thisaddress had the
2981 value of $0, $0 had the value of $1, etc). This bug was introduced by
2982 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2984 38. Improved the line breaking for long SMTP error messages from ACLs.
2985 Previously, if there was no break point between 40 and 75 characters, Exim
2986 left the rest of the message alone. Two changes have been made: (a) I've
2987 reduced the minimum length to 35 characters; (b) if it can't find a break
2988 point between 35 and 75 characters, it looks ahead and uses the first one
2989 that it finds. This may give the occasional overlong line, but at least the
2990 remaining text gets split now.
2992 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2993 file descriptors might be low, and that setting 1000 would always raise it.
2994 It turns out that in some environments, the limit is already over 1000 and
2995 that lowering it causes trouble. So now Exim takes care not to decrease it.
2997 40. When delivering a message, the value of $return_path is set to $sender_
2998 address at the start of routing (routers may change the value). By an
2999 oversight, this default was not being set up when an address was tested by
3000 -bt or -bv, which affected the outcome if any router or filter referred to
3003 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3004 log only when "message" or "log_message" are set. However, if one of the
3005 conditions was an address verification, or a call to a nested ACL, the
3006 messages generated by the underlying test were being passed through. This
3007 no longer happens. The underlying message is available in $acl_verify_
3008 message for both "message" and "log_message" expansions, so it can be
3009 passed through if needed.
3011 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3012 new expansion $bh_ to give the encoded byte string without charset
3013 translation. Translation happens only if iconv() is available; HAVE_ICONV
3014 indicates this at build time. HEADERS_CHARSET gives the charset to
3015 translate to; headers_charset can change it in the configuration, and
3016 "headers charset" can change it in an individual filter file.
3018 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3019 that creates RFC 2047 encoded "words" labels them as that charset instead
3020 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3021 expansion operator; (ii) when Exim creates a From: line for a local
3022 message; (iii) when a header line is rewritten to include a "phrase" part.
3024 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3025 buggy, causing it to skip the first lines of messages whose message ID
3026 ended in 'D'. This would not have bitten before Exim release 4.14, because
3027 message IDs were unlikely to end in 'D' before then. The effect was to have
3028 incorrect size information for certain domains.
3030 45. #include "config.h" was missing at the start of the crypt16.c module. This
3031 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3034 46. If there was a timeout during a "random" callout check, Exim treated it as
3035 a failure of the random address, and carried on sending RSET and the real
3036 address. If the delay was just some slowness somewhere, the response to the
3037 original RCPT would be taken as a response to RSET and so on, causing
3038 mayhem of various kinds.
3040 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3041 when I implemented it. It didn't allow for the fact that some option values
3042 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3043 the right test for positive values.
3045 48. Domain names in DNS records are case-independent. Exim always looks them up
3046 in lower case. Some resolvers return domain names in exactly the case they
3047 appear in the zone file, that is, they may contain uppercase letters. Not
3048 all resolvers do this - some return always lower case. Exim was treating a
3049 change of case by a resolver as a change of domain, similar to a widening
3050 of a domain abbreviation. This triggered its re-routing code and so it was
3051 trying to route what was effectively the same domain again. This normally
3052 caused routing to fail (because the router wouldn't handle the domain
3053 twice). Now Exim checks for this case specially, and just changes the
3054 casing of the domain that it ultimately uses when it transmits the message
3057 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3060 50. If a filter generated a file delivery with a non-absolute name (possible if
3061 no home directory exists for the router), the forbid_file option was not
3064 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3065 the existing equality matching.
3067 52. Exim was using ints instead of ino_t variables in some places where it was
3068 dealing with inode numbers.
3070 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3071 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3072 called TMPDIR, and if it finds it is different, it changes its value.
3074 54. The smtp_printf() function is now made available to local_scan() so
3075 additional output lines can be written before returning. There is also an
3076 smtp_fflush() function to enable the detection of a dropped connection.
3077 The variables smtp_input and smtp_batched_input are exported to
3080 55. Changed the default runtime configuration: the message "Unknown user"
3081 has been removed from the ACL, and instead placed on the localuser router,
3082 using the cannot_route_message feature. This means that any verification
3083 failures that generate their own messages won't get overridden. Similarly,
3084 the "Unrouteable address" message that was in the ACL for unverifiable
3085 relay addresses has also been removed.
3087 56. Added hosts_avoid_esmtp to the smtp transport.
3089 57. The exicyclog script was not checking for the esoteric option
3090 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3091 will work only if exicyclog is run under the appropriate euid.
3093 58. Following a discussion on the list, the rules by which Exim recognises line
3094 endings on incoming messages have been changed. The -dropcr and drop_cr
3095 options are now no-ops, retained only for backwards compatibility. The
3096 following line terminators are recognized: LF CRLF CR. However, special
3097 processing applies to CR:
3099 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3100 nor a local message in the state where . is a terminator.
3102 (ii) If a bare CR is encountered in a header line, an extra space is added
3103 after the line terminator so as not to end the header. The reasoning
3104 behind this is that bare CRs in header lines are most likely either
3105 to be mistakes, or people trying to play silly games.
3107 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3108 was being incorrectly given as 18 bytes larger than it should have been.
3109 This is a VOB (very old bug).
3111 60. This may never have affected anything current, but just in case it has:
3112 When the local host is found other than at the start of a list of hosts,
3113 the local host, those with the same MX, and any that follow, are discarded.
3114 When the list in question was part of a longer list of hosts, the following
3115 hosts (not currently being processed) were also being discarded. This no
3116 longer happens. I'm not sure if this situation could ever has previously
3119 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3122 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3123 header. This is something that is recommended in a new Internet Draft, and
3124 is something that is documented as being done by Sendmail. There are two
3125 possible values. For messages generated by the autoreply transport, Exim
3128 Auto-Submitted: auto-replied
3130 whereas for all other generated messages (e.g. bounces) it adds
3132 Auto-Submitted: auto-generated
3134 63. The "personal" condition in filters now includes a test for the
3135 Auto-Submitted: header. If it contains the string "auto-" the message it
3136 not considered personal.
3138 64. Added rcpt_include_affixes as a generic transport option.
3140 65. Added queue_only_override (default true).
3142 66. Added the syslog_duplication option.
3144 67. If what should have been the first header line of a message consisted of
3145 a space followed by a colon, Exim was mis-interpreting it as a header line.
3146 It isn't of course - it is syntactically invalid and should therefore be
3147 treated as the start of the message body. The misbehaviour could have
3148 caused a number of strange effects, including loss of data in subsequent
3149 header lines, and spool format errors.
3151 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3152 client host had authenticated. This control can now be exercised by an ACL
3153 for more flexibility.
3155 69. By default, callouts do not happen when testing with -bh. There is now a
3156 variant, -bhc, which does actually run the callout code, including
3157 consulting and updating the callout cache.
3159 70. Added support for saslauthd authentication, courtesy of Alexander
3162 71. If statvfs() failed on the spool or log directories while checking their
3163 size for availability, Exim confusingly gave the error "space shortage".
3164 Furthermore, in debugging mode it crashed with a floating point exception.
3165 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3166 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3167 serious problem, Exim now writes to the main and panic logs when this
3168 happens, with details of the failure. It then refuses to accept the
3169 incoming message, giving the message "spool directory problem" or "log
3170 directory problem" with a 421 code for SMTP messages.
3172 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3173 0, 1, and 2 exist, because some OS complain for execs without them (see
3174 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3175 descriptors. However, the code omitted to check that the open succeeded,
3176 causing mysterious errors if for some reason the permissions on /dev/null
3177 got screwed. Now Exim writes a message to the main and panic logs, and
3178 bombs out if it can't open /dev/null.
3180 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3181 interact so that it is all more flexible. It is supposed to remain
3182 backwards compatible. Also added extra_local_interfaces.
3184 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3185 to bomb out with an assertion failure - to the client this appears as a
3186 connection drop. This problem occurs in the part of the code that was taken
3187 from the Samba project. Fortunately, the assertion is in a very simple
3188 function, so I have fixed this by reproducing the function inline in the
3189 one place where it is called, and arranging for authentication to fail
3190 instead of killing the process with assert().
3192 75. The SPA client code was not working when the server requested OEM rather
3193 than Unicode encoding.
3195 76. Added code to make require_files with a specific uid setting more usable in
3196 the case where statting the file as root fails - usually a non-root-mounted
3197 NFS file system. When this happens and the failure is EACCES, Exim now
3198 forks a subprocess and does the per-uid checking as the relevant uid.
3200 77. Added process_log_path.
3202 78. If log_file_path was not explicitly set, a setting of check_log_space or
3203 check_log_inodes was ignored.
3205 79. If a space check for the spool or log partitions fails, the incident is now
3206 logged. Of course, in the latter case the data may get lost...
3208 80. Added the %p formatting code to string_format() so that it can be used to
3209 print addresses in debug_print(). Adjusted all the address printing in the
3210 debugging in store.c to use %p rather than %d.
3212 81. There was a concern that a line of code in smtp_in.c could overflow a
3213 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3214 initially expressed, the concern was not well-founded, because trailing
3215 spaces are removed early. However, if the trailing spaces were followed by
3216 a NULL, they did not get removed, so the overflow was possible. Two fixes
3219 (a) I re-wrote the offending code in a cleaner fashion.
3220 (b) If an incoming SMTP command contains a NULL character, it is rejected
3223 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3224 runs initgroups(), so that if the Exim user is in any additional groups,
3225 they will be used during message reception.
3231 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob