1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.195 2005/08/02 13:43:04 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.
57 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
59 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
60 command itself ended in error, the underlying message about the transport
61 filter timeout was being overwritten with the pipe command error. Now the
62 underlying error message should be appended to the second error message.
64 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
66 PH/02 Reduce the amount of output that "make" produces by default. Full output
67 can still be requested.
69 PH/03 The warning log line about a condition test deferring for a "warn" verb
70 was being output only once per connection, rather than after each
71 occurrence (because it was using the same function as for successful
72 "warn" verbs). This seems wrong, so I have changed it.
74 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
75 it should not have, which might have caused a crash in the right
76 circumstances, but probably never did.
78 PH/04 Installed a modified version of Tony Finch's patch to make submission
79 mode fix the return path as well as the Sender: header line, and to
80 add a /name= option so that you can make the user's friendly name appear
83 TF/03 Added the control = fakedefer ACL modifier.
85 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
86 Mark Lowes for thorough testing.
88 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
90 TK/03 Merged latest SRS patch from Miles Wilton.
92 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
93 with the definition in sysexits.h (which is #included earlier).
94 Fortunately, Exim does not actually use EX_OK. The code used to try to
95 preserve the sysexits.h value, by assumimg that macro definitions were
96 scanned for macro replacements. I have been disabused of this notion,
97 so now the code just undefines EX_OK before #including unistd.h.
99 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
100 in the smtp transport. When a block could not be written in a single
101 write() function, the timeout was being re-applied to each part-write.
102 This seems wrong - if the receiver was accepting one byte at a time it
103 would take for ever. The timeout is now adjusted when this happens. It
104 doesn't have to be particularly precise.
106 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
107 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
109 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
110 without a domain if the check to <postmaster@domain> fails.
112 SC/01 Eximstats: added -xls and the ability to specify output files
113 (patch written by Frank Heydlauf).
115 SC/02 Eximstats: use FileHandles for outputing results.
117 SC/03 Eximstats: allow any combination of xls, txt, and html output.
119 SC/04 Eximstats: fixed display of large numbers with -nvr option
121 SC/05 Eximstats: fixed merging of reports with empty tables.
123 SC/06 Eximstats: added the -include_original_destination flag
125 SC/07 Eximstats: removed tabs and trailing whitespace.
127 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
129 TK/06 MBOX spool code: Add real "From " MBOX separator line
130 so the .eml file is really in mbox format (even though
131 most programs do not really care). Patch from Alex Miller.
133 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
134 The latter is generated from $received_to and is only set if the
135 message has one envelope recipient. SA can use these headers,
136 obviously out-of-the-box. Patch from Alex Miller.
138 PH/08 The ${def test on a variable was returning false if the variable's
139 value was "0", contrary to what the specification has always said!
140 The result should be true unless the variable is empty.
142 PH/09 The syntax error of a character other than { following "${if
143 def:variable_name" (after optional whitespace) was not being diagnosed.
144 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
145 accidental colon was present, for example, could give incorrect results.
147 PH/10 Tidied the code in a number of places where the st_size field of a stat()
148 result is used (not including appendfile, where other changes are about
151 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
152 This involved changing a lot of size variables from int to off_t. It
153 should work with maildirs and everything.
155 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
156 spamd dying while we are connected to it.
158 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
161 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
162 function for BDB 4.3.
164 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
166 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
167 the sources for PCRE have been re-arranged and more files are now
170 PH/15 The code I had for printing potentially long long variables in PH/11
171 above was not the best (it lost precision). The length of off_t variables
172 is now inspected at build time, and an appropriate printing format (%ld
173 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
174 to be "long long int" or "long int". This is needed for the internal
175 formatting function string_vformat().
177 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
178 the configuration file to be ":syslog", then the script "guesses" where
179 the logs files are, rather than using the compiled in default. In our
180 case the guess is not the same as the compiled default, so the script
181 suddenly stopped working when I started to use syslog. The patch checks
182 to see if log_file_path is "". If so, it attempts to read it from exim
183 with no configuration file to get the compiled in version, before it
184 falls back to the previous guessing code."
186 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
187 implementing BATV in an Exim configuration. See NewStuff for the gory
190 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
191 Makefile that are specific to HP-UX.
193 PH/18 If the "use_postmaster" option was set for a recipient callout together
194 with the "random" option, the postmaster address was used as the MAIL
195 FROM address for the random test, but not for the subsequent recipient
196 test. It is now used for both.
198 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
199 patch removes a few documentation additions to RFC 3028, because the
200 latest draft now contains them. It adds the new en;ascii-case comparator
201 and a new error check for 8bit text in MIME parts. Comparator and
202 require names are now matched exactly. I enabled the subaddress
203 extension, but it is not well tested yet (read: it works for me)."
205 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
206 rework some of the code of TK/09 above to avoid the hardwired use of
207 "%lld" and "long long". Replaced the call to snprintf() with a call to
210 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
211 records point to non-existent hosts", "retry timeout exceeded", and
212 "retry time not reached for any host after a long failure period".
214 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
215 experimental DomainKeys support:
217 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
218 (2) On an error such as an illegally used "control", the wrong name for
219 the control was given.
221 These problems did NOT occur unless DomainKeys support was compiled.
223 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
225 PH/24 Added ${if match_ip condition.
227 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
228 databases so that it will be absolutely obvious if a crash occurs in the
229 DB library. This is a regular occurrence (often caused by mis-matched
232 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
233 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
234 picked up on a user's system that detects such things. There doesn't seem
235 to be a gcc warning option for this - only an attribute that has to be
236 put on the function's prototype. It seems that in Fedora Core 4 they have
237 set this on a number of new functions. No doubt there will be more in due
240 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
241 specify a transport. However, if an address that was verified by such a
242 router was the subject of a callout, Exim crashed because it tried to
243 read the rcpt_include_affixes from the non-existent transport. Now it
244 just assumes that the setting of that option is false. This bug was
245 introduced by 4.51/PH/31.
247 PH/28 Changed -d+all to exclude +memory, because that information is very
248 rarely of interest, but it makes the output a lot bigger. People tend to
249 do -d+all out of habit.
251 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
252 code in os-type was giving problems when libc.so lives in lib64, like on
255 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
256 aren't the modern standard, and it seems that some systems' include files
257 don't always have them. Exim was already checking for some of the newer
258 ones like T_AAAA, and defining it itself. I've added checks for all the
259 record types that Exim uses.
261 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
262 not automatically generating a new one, as it is supposed to. This
263 prevented TLS from working. If the file did exist, but contained invalid
264 data, a new version was generated, as expected. It was only the case of a
265 non-existent file that was broken.
267 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
268 with a change in libdomainkeys > 0.64.
270 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
271 from DNS. If the selector record carries the flag, it now has
272 precedence over the domain-wide flag.
274 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
276 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
277 the use of a port name that isn't defined in /etc/services) occurred, the
278 message was deferred as in a normal delivery, and thus remained on the
279 spool, instead of being failed because of the mua_wrapper setting. This
280 is now fixed, and I tidied up some of the mua_wrapper messages at the
283 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
284 the messages in a hash of arrays rather than using individual hashes.
285 This is a bit cleaner and results in dramatic memory savings, albeit
286 at a slight CPU cost.
288 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
289 as requested by Marc Sherman.
291 SC/10 Eximstats: added histograms for user specified patterns as requested
294 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
296 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
297 fopen() in the content-scanning modules that did not already have it.
303 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
304 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
306 TK/02 Fix ACL "control" statement not being available in MIME ACL.
308 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
310 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
311 to test Sieve filters that use "vacation".
313 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
314 that changes the way the GnuTLS parameters are stored in the cache file.
315 The new format can be generated externally. For backward compatibility,
316 if the data in the cache doesn't make sense, Exim assumes it has read an
317 old-format file, and it generates new data and writes a new file. This
318 means that you can't go back to an older release without removing the
321 PH/03 A redirect router that has both "unseen" and "one_time" set does not
322 work if there are any delivery delays because "one_time" forces the
323 parent to be marked "delivered", so its unseen clone is never tried
324 again. For this reason, Exim now forbids the simultaneous setting of
327 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
328 redirected to themselves ("homonym" addresses). Read the long ChangeLog
329 entry if you want to know the details. The fix, however, neglected to
330 consider the case when local delivery batching is involved. The test for
331 "previously delivered" was not happening when checking to see if an
332 address could be batched with a previous (undelivered) one; under
333 certain circumstances this could lead to multiple deliveries to the same
336 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
337 in its include files, and this causes problems building Exim.
339 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
340 header_syntax) but Exim was just ignoring anything given after a slash.
341 In particular, this caused confusion with an attempt to use "verify =
342 reverse_host_lookup/defer_ok". An error is now given when options are
343 supplied for verify items that do not have them. (Maybe reverse_host_
344 lookup should have a defer_ok option, but that's a different point.)
346 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
347 defined by RFC 821) to 2048, because there were problems with some AUTH
348 commands, and RFC 1869 says the size should be increased for extended
349 SMTP commands that take arguments.
351 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
354 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
355 "unknown" error; now it says that the functionality isn't in the binary.
357 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
358 an address' error message when a string expansion fails (syntax or
359 whatever). Otherwise the password may appear in the log. Following change
360 PH/42 below, there is no longer a chance of it appearing in a bounce
363 PH/11 Installed exipick version 20050225.0 from John Jetmore.
365 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
366 of its addresses was ever tried. (Bugzilla bug #2.)
368 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
369 the result incorrectly in the debug output. (It correctly added a newline
370 to what was transported.)
372 TF/01 Added $received_time.
374 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
375 commented out examples of how to interface to a virus scanner and to
376 SpamAssassin. Also added commented examples of av_scanner and
377 spamd_address settings.
379 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
380 and controls are allowed in which ACLs. There were a couple of minor
381 errors. Some of the entries in the conditions table (which is a table of
382 where they are NOT allowed) were getting very unwieldy; rewrote them as a
383 negation of where the condition IS allowed.
385 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
387 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
388 header file does not have a version number, so I've had to invent a new
389 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
390 API. The code is untested by me (my Linux distribution still has 0.3.2 of
391 radiusclient), but it was contributed by a Radius user.
393 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
394 files or directories, for OpenSSL.
396 PH/19 When an Exim process that is running as root has to create an Exim log
397 file, it does so in a subprocess that runs as exim:exim so as to get the
398 ownership right at creation (otherwise, other Exim processes might see
399 the file with the wrong ownership). There was no test for failure of this
400 fork() call, which would lead to the process getting stuck as it waited
401 for a non-existent subprocess. Forks do occasionally fail when resources
402 run out. I reviewed all the other calls to fork(); they all seem to check
405 PH/20 When checking for unexpected SMTP input at connect time (before writing
406 the banner), Exim was not dealing correctly with a non-positive return
407 from the read() function. If the client had disconnected by this time,
408 the result was a log entry for a synchronization error with an empty
409 string after "input=" when read() returned zero. If read() returned -1
410 (an event I could not check), uninitialized data bytes were printed.
411 There were reports of junk text (parts of files, etc) appearing after
414 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
416 PH/22 Added support for macro redefinition, and (re)definition in between
417 driver and ACL definitions.
419 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
420 forgetting to use the resulting value; it was using the unexpanded value.
422 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
423 hadn't been configured. The fix is from Juergen Kreileder, who
424 understands it better than I do:
426 "Here's what I see happening with three configured cyrus_sasl
427 authenticators configured (plain, login, cram-md5):
429 On startup auth_cyrus_sasl_init() gets called for each of these.
430 This means three calls to sasl_listmech() without a specified mech_list.
431 => SASL tests which mechs of all available mechs actually work
432 => three warnings about OTP not working
433 => the returned list contains: plain, login, cram-md5, digest-md5, ...
435 With the patch, sasl_listmech() also gets called three times. But now
436 SASL's mech_list option is set to the server_mech specified in the the
437 authenticator. Or in other words, the answer from sasl_listmech()
438 gets limited to just the mech you're testing for (which is different
440 => the return list contains just 'plain' or 'login', 'cram-md5' or
441 nothing depending on the value of ob->server_mech.
443 I've just tested the patch: Authentication still works fine,
444 unavailable mechs specified in the exim configuration are still
445 caught, and the auth.log warnings about OTP are gone."
447 PH/25 When debugging is enabled, the contents of the command line are added
448 to the debugging output, even when log_selector=+arguments is not
451 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
452 answer is "GNU", and only if the return is "GNU/something" is the answer
455 PH/27 $acl_verify_message is now set immediately after the failure of a
456 verification in an ACL, and so is available in subsequent modifiers. In
457 particular, the message can be preserved by coding like this:
459 warn !verify = sender
460 set acl_m0 = $acl_verify_message
462 Previously, $acl_verify_message was set only while expanding "message"
463 and "log_message" when a very denied access.
465 PH/28 Modified OS/os.c-Linux with
467 -#ifndef OS_LOAD_AVERAGE
468 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
470 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
471 nomenclature these days.)
473 PH/29 Installed patch from the Sieve maintainer that adds the options
474 sieve_useraddress and sieve_subaddress to the redirect router.
476 PH/30 In these circumstances:
477 . Two addresses routed to the same list of hosts;
478 . First host does not offer TLS;
479 . First host accepts first address;
480 . First host gives temporary error to second address;
481 . Second host offers TLS and a TLS session is established;
482 . Second host accepts second address.
483 Exim incorrectly logged both deliveries with the TLS parameters (cipher
484 and peerdn, if requested) that were in fact used only for the second
487 PH/31 When doing a callout as part of verifying an address, Exim was not paying
488 attention to any local part prefix or suffix that was matched by the
489 router that accepted the address. It now behaves in the same way as it
490 does for delivery: the affixes are removed from the local part unless
491 rcpt_include_affixes is set on the transport.
493 PH/32 Add the sender address, as F=<...>, to the log line when logging a
494 timeout during the DATA phase of an incoming message.
496 PH/33 Sieve envelope tests were broken for match types other than :is. I have
497 applied a patch sanctioned by the Sieve maintainer.
499 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
500 the uid or gid is negative. A case of a negative gid caused this to be
501 noticed. The fix allows for either to be negative.
503 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
504 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
507 PH/36 The change PH/12 above was broken. Fixed it.
509 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
510 the grounds that routing the same address twice would always produce the
511 same answer. This might have been true once, but it is certainly no
512 longer true now. Routing a child address may depend on the previous
513 routing that produced that child. Some complicated redirection strategies
514 went wrong when messages had multiple recipients, and made Exim's
515 behaviour dependent on the order in which the addresses were given.
517 I have moved the duplicate checking until after the routing is complete.
518 Exim scans the addresses that are assigned to local and remote
519 transports, and removes any duplicates. This means that more work will be
520 done, as duplicates will always all be routed, but duplicates are
521 presumably rare, so I don't expect this is of any significance.
523 For deliveries to pipes, files, and autoreplies, the duplicate checking
524 still happens during the routing process, since they are not going to be
527 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
528 It corrects a timeout issue with spamd. This is Ian's comment: "The
529 background is that sometimes spamd either never reads data from a
530 connection it has accepted, or it never writes response data. The exiscan
531 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
532 blindly assumes that writes won't block so it may never time out."
534 PH/39 Allow G after quota size as well as K and M.
536 PH/40 The value set for $authenticated_id in an authenticator may not contain
537 binary zeroes or newlines because the value is written to log lines and
538 to spool files. There was no check on this. Now the value is run through
539 the string_printing() function so that such characters are converted to
540 printable escape sequences.
542 PH/41 $message_linecount is a new variable that contains the total number of
543 lines in the message. Compare $body_linecount, which is the count for the
546 PH/42 Exim no longer gives details of delivery errors for specific addresses in
547 bounce and delay warning messages, except in certain special cases, which
550 (a) An SMTP error message from a remote host;
551 (b) A message specified in a :fail: redirection;
552 (c) A message specified in a "fail" command in a system filter;
553 (d) A message specified in a FAIL return from the queryprogram router;
554 (e) A message specified by the cannot_route_message router option.
556 In these cases only, Exim does include the error details in bounce and
557 warning messages. There are also a few cases where bland messages such
558 as "unrouteable address" or "local delivery error" are given.
560 PH/43 $value is now also set for the "else" part of a ${run expansion.
562 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
563 being worked on, but at least Exim now implements the latest version to
566 PH/45 In a pipe transport, although a timeout while waiting for the pipe
567 process to complete was treated as a delivery failure, a timeout while
568 writing the message to the pipe was logged, but erroneously treated as a
569 successful delivery. Such timeouts include transport filter timeouts. For
570 consistency with the overall process timeout, these timeouts are now
571 treated as errors, giving rise to delivery failures by default. However,
572 there is now a new Boolean option for the pipe transport called
573 timeout_defer, which, if set TRUE, converts the failures into defers for
574 both kinds of timeout. A transport filter timeout is now identified in
577 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
578 systems where "make" and "gmake" are different, calling "gmake" at top
579 level broke things. I've arranged for the value of $(MAKE) to be passed
580 from the Makefile to this script so that it can call the same version of
584 A note about Exim versions 4.44 and 4.50
585 ----------------------------------------
587 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
588 changes of various kinds. As a consequence, a big documentation update was
589 needed. This delayed the release for rather longer than seemed good, especially
590 in the light of a couple of (minor) security issues. Therefore, the changes
591 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
592 release. So 4.44 and 4.50 are in effect two different branches that both start
595 I have left the 4.50 change log unchanged; it contains all the changes since
596 4.43. The change log for 4.44 is below; many of its items are identical to
597 those for 4.50. This seems to be the most sensible way to preserve the
598 historical information.
604 1. Minor wording change to the doc/README.SIEVE file.
606 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
607 computation of the current number of files was incorrect.
609 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
610 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
613 4. Give more explanation in the error message when the command for a transport
614 filter fails to execute.
616 5. There are several places where Exim runs a non-Exim command in a
617 subprocess. The SIGUSR1 signal should be disabled for these processes. This
618 was being done only for the command run by the queryprogram router. It is
619 now done for all such subprocesses. The other cases are: ${run, transport
620 filters, and the commands run by the lmtp and pipe transports.
622 6. Added CONFIGURE_GROUP build-time option.
624 7. Some older OS have a limit of 256 on the maximum number of file
625 descriptors. Exim was using setrlimit() to set 1000 as a large value
626 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
627 systems. I've change it so that if it can't get 1000, it tries for 256.
629 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
630 was an oversight, and furthermore, ever since the addition of extra
631 controls (e.g. 4.43/32), the checks on when to allow different forms of
632 "control" were broken. There should now be diagnostics for all cases when a
633 control that does not make sense is encountered.
635 9. Added the /retain_sender option to "control=submission".
637 10. $recipients is now available in the predata ACL (oversight).
639 11. Tidy the search cache before the fork to do a delivery from a message
640 received from the command line. Otherwise the child will trigger a lookup
641 failure and thereby defer the delivery if it tries to use (for example) a
642 cached ldap connection that the parent has called unbind on.
644 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
645 of $address_data from the recipient verification was clobbered by the
648 13. The value of address_data from a sender verification is now available in
649 $sender_address_data in subsequent conditions in the ACL statement.
651 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
653 15. Added a new option "connect=<time>" to callout options, to set a different
656 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
657 was its contents. (It was OK if the option was not defined at all.)
659 17. A "Completed" log line is now written for messages that are removed from
660 the spool by the -Mrm option.
662 18. New variables $sender_verify_failure and $recipient_verify_failure contain
663 information about exactly what failed.
665 19. Added -dd to debug only the daemon process.
667 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
668 handles timeouts, both on the server side and network timeouts. Renamed the
669 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
671 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
672 It is now set to "smtps".
674 22. $host_address is now set to the target address during the checking of
677 23. When checking ignore_target_hosts for an ipliteral router, no host name was
678 being passed; this would have caused $sender_host_name to have been used if
679 matching the list had actually called for a host name (not very likely,
680 since this list is usually IP addresses). A host name is now passed as
683 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
684 code that specifies a non-restarting handler (typically sigaction() in
685 modern systems) in an attempt to fix a rare and obscure crash bug.
687 25. Narrowed the window for a race in the daemon that could cause it to ignore
688 SIGCHLD signals. This is not a major problem, because they are used only to
689 wake it up if nothing else does.
691 26. A malformed maildirsize file could cause Exim to calculate negative values
692 for the mailbox size or file count. Odd effects could occur as a result.
693 The maildirsize information is now recalculated if the size or filecount
696 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
697 support for a long time. Removed HAVE_SYS_VFS_H.
699 28. Installed the latest version of exipick from John Jetmore.
701 29. In an address list, if the pattern was not a regular expression, an empty
702 subject address (from a bounce message) matched only if the pattern was an
703 empty string. Non-empty patterns were not even tested. This was the wrong
704 because it is perfectly reasonable to use an empty address as part of a
705 database query. An empty address is now tested by patterns that are
706 lookups. However, all the other forms of pattern expect the subject to
707 contain a local part and a domain, and therefore, for them, an empty
708 address still always fails if the pattern is not itself empty.
710 30. Exim went into a mad DNS loop when attempting to do a callout where the
711 host was specified on an smtp transport, and looking it up yielded more
714 31. Re-factored the code for checking spool and log partition space into a
715 function that finds that data and another that does the check. The former
716 is then used to implement four new variables: $spool_space, $log_space,
717 $spool_inodes, and $log_inodes.
719 32. The RFC2047 encoding function was originally intended for short strings
720 such as real names; it was not keeping to the 75-character limit for
721 encoded words that the RFC imposes. It now respects the limit, and
722 generates multiple encoded words if necessary. To be on the safe side, I
723 have increased the buffer size for the ${rfc2047: expansion operator from
726 33. It is now permitted to omit both strings after an "if" condition; if the
727 condition is true, the result is "true". As before, when the second string
728 is omitted, a false condition yields an empty string. This makes it less
729 cumbersome to write custom ACL and router conditions.
731 34. Failure to deliver a bounce message always caused it to be frozen, even if
732 there was an errors_to setting on the router. The errors_to setting is now
735 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
736 canonical form (fully expanded) before being placed in
737 $sender_host_address.
739 36. The table in the code that translates DNS record types into text (T_A to
740 "A" for instance) was missing entries for NS and CNAME. It is just possible
741 that this could have caused confusion if both these types were looked up
742 for the same domain, because the text type is used as part of Exim's
743 per-process caching. But the chance of anyone hitting this buglet seems
746 37. The dnsdb lookup has been extended in a number of ways.
748 (1) There is a new type, "zns", which walks up the domain tree until it
749 finds some nameserver records. It should be used with care.
751 (2) There is a new type, "mxh", which is like "mx" except that it returns
752 just the host names, not the priorities.
754 (3) It is now possible to give a list of domains (or IP addresses) to be
755 looked up. The behaviour when one of the lookups defers can be
756 controlled by a keyword.
758 (4) It is now possible to specify the separator character for use when
759 multiple records are returned.
761 38. The dnslists ACL condition has been extended: it is now possible to supply
762 a list of IP addresses and/or domains to be looked up in a particular DNS
765 39. Added log_selector=+queue_time_overall.
767 40. When running the queue in the test harness, wait just a tad after forking a
768 delivery process, to get repeatability of debugging output.
770 41. Include certificate and key file names in error message when GnuTLS fails
771 to set them up, because the GnuTLS error message doesn't include the name
772 of the failing file when there is a problem reading it.
774 42. Allow both -bf and -bF in the same test run.
776 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
778 44. The "Exiscan patch" is now merged into the mainline Exim source.
780 45. Sometimes the final signoff response after QUIT could fail to get
781 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
782 < 0 before doing a fflush(). This bug looks as though it goes back to the
783 introduction of TLS in release 3.20, but "sometimes" must have been rare
784 because the tests only now provoked it.
786 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
787 (this can affect the format of dates).
789 47. exim_tidydb, when checking for the continued existence of a message for
790 which it has found a message-specific retry record, was not finding
791 messages that were in split spool directories. Consequently, it was
792 deleting retry records that should have stayed in existence.
794 48. Steve fixed some bugs in eximstats.
796 49. The SPA authentication driver was not abandoning authentication and moving
797 on to the next authenticator when an expansion was forced to fail,
798 contradicting the general specification for all authenticators. Instead it
799 was generating a temporary error. It now behaves as specified.
801 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
802 (the order specifies the preference for clients). The order is now AES256,
803 AES128, 3DES, ARCFOUR128.
805 51. Small patch to Sieve code - explicitly set From: when generating an
808 52. Exim crashed if a remote delivery caused a very long error message to be
809 recorded - for instance if somebody sent an entire SpamAssassin report back
810 as a large number of 550 error lines. This bug was coincidentally fixed by
811 increasing the size of one of Exim's internal buffers (big_buffer) that
812 happened as part of the Exiscan merge. However, to be on the safe side, I
813 have made the code more robust (and fixed the comments that describe what
816 53. Now that there can be additional text after "Completed" in log lines (if
817 the queue_time_overall log selector is set), a one-byte patch to exigrep
818 was needed to allow it to recognize "Completed" as not the last thing in
821 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
822 patch that reportedly fixes this has been added. I am not expert enough to
823 create a test for it. This is what the patch creator wrote:
825 "I found a little strange behaviour of ldap code when working with
826 Windows 2003 AD Domain, where users was placed in more than one
827 Organization Units. When I tried to give exim partial DN, the exit code
828 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
829 But simultaneously result of request was absolutely normal ldap result,
830 so I produce this patch..."
832 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
833 REFERENCE, so I have modified the code to exclude the patch when that macro
836 55. Some experimental protocols are using DNS PTR records for new purposes. The
837 keys for these records are domain names, not reversed IP addresses. The
838 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
839 leaves it alone. Component reversal etc. now happens only for IP addresses.
842 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
844 57. Double the size of the debug message buffer (to 2048) so that more of very
845 long debug lines gets shown.
847 58. The exicyclog utility now does better if the number of log files to keep
848 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
850 59. Two changes related to the smtp_active_hostname option:
852 (1) $smtp_active_hostname is now available as a variable.
853 (2) The default for smtp_banner uses $smtp_active_hostname instead
854 of $primary_hostname.
856 60. The host_aton() function is supposed to be passed a string that is known
857 to be a valid IP address. However, in the case of IPv6 addresses, it was
858 not checking this. This is a hostage to fortune. Exim now panics and dies
859 if the condition is not met. A case was found where this could be provoked
860 from a dnsdb PTR lookup with an IPv6 address that had more than 8
861 components; fortuitously, this particular loophole had already been fixed
862 by change 4.50/55 above.
864 If there are any other similar loopholes, the new check in host_aton()
865 itself should stop them being exploited. The report I received stated that
866 data on the command line could provoke the exploit when Exim was running as
867 exim, but did not say which command line option was involved. All I could
868 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
872 61. There was a buffer overflow vulnerability in the SPA authentication code
873 (which came originally from the Samba project). I have added a test to the
874 spa_base64_to_bits() function which I hope fixes it.
877 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
878 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
880 63. The daemon start-up calls getloadavg() while still root for those OS that
881 need the first call to be done as root, but it missed one case: when
882 deliver_queue_load_max is set with deliver_drop_privilege. This is
883 necessary for the benefit of the queue runner, because there is no re-exec
884 when deliver_drop_privilege is set.
886 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
889 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
890 was in use, was not putting the data itself into the right store pool;
891 consequently, it could be overwritten for a subsequent message in the same
892 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
895 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
897 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
898 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
899 the function were treating the return as a boolean value, which happened to
900 work because 0=false and not-0=true, but is not correct code.
902 68. The host_aton() function was not handling scoped IPv6 addresses (those
903 with, for example, "%eth0" on the end) correctly.
905 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
906 negated items (that is, ~something) in unsigned ints. Some compilers
907 apparently mutter when there is no cast.
909 70. If an address verification called from an ACL failed, and did not produce a
910 user-specific message (i.e. there was only a "system" message), nothing was
911 put in $acl_verify_message. In this situation, it now puts the system
914 71. Change 4.23/11 added synchronization checking at the start of an SMTP
915 session; change 4.31/43 added the unwanted input to the log line - except
916 that it did not do this in the start of session case. It now does.
918 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
919 This is wrong and can cause the other end to generate a synchronization
920 error if it is another Exim or anything else that does the synchronization
921 check. A QUIT command is no longer sent after a timeout.
923 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
926 74. The defer_ok option of callout verification was not working if it was used
927 when verifying addresses in header lines, that is, for this case:
929 verify = header_sender/callout=defer_ok
931 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
932 those file descriptors could be used for SMTP connections. If anything
933 wrote to stderr (the example that came up was "warn" in embedded Perl), it
934 could be sent to the SMTP client, causing chaos. The daemon now opens
935 stdin, stdout, and stderr to /dev/null when it puts itself into the
938 76. Arrange for output from Perl's "warn" command to be written to Exim's main
939 log by default. The user can override this with suitable Perl magic.
941 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
942 the log message when discard triggers, was not working for the DATA ACL or
943 for the non-SMTP ACL.
945 78. Error message wording change in sieve.c.
947 79. If smtp_accept_max_per_host was set, the number of connections could be
948 restricted to fewer than expected, because the daemon was trying to set up
949 a new connection before checking whether the processes handling previous
950 connections had finished. The check for completed processes is now done
951 earlier. On busy systems, this bug wouldn't be noticed because something
952 else would have woken the daemon, and it would have reaped the completed
955 80. If a message was submitted locally by a user whose login name contained one
956 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
957 It caused a spool format error. I have fixed the spool reading code. A
958 related problem was that the "from" clause in the Received: line became
959 illegal because of the space(s). It is now covered by ${quote_local_part.
961 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
964 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
966 83. Patch from Sieve maintainer to fix unterminated string problem in
969 84. Some minor changes to the Linux configuration files to help with other
970 OS variants using glibc.
972 85. One more patch for Sieve to update vacation handling to latest spec.
975 ----------------------------------------------------
976 See the note above about the 4.44 and 4.50 releases.
977 ----------------------------------------------------
983 1. Change 4.43/35 introduced a bug that caused file counts to be
984 incorrectly computed when quota_filecount was set in an appendfile
987 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
988 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
991 3. Give more explanation in the error message when the command for a transport
992 filter fails to execute.
994 4. There are several places where Exim runs a non-Exim command in a
995 subprocess. The SIGUSR1 signal should be disabled for these processes. This
996 was being done only for the command run by the queryprogram router. It is
997 now done for all such subprocesses. The other cases are: ${run, transport
998 filters, and the commands run by the lmtp and pipe transports.
1000 5. Some older OS have a limit of 256 on the maximum number of file
1001 descriptors. Exim was using setrlimit() to set 1000 as a large value
1002 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1003 systems. I've change it so that if it can't get 1000, it tries for 256.
1005 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1006 was an oversight, and furthermore, ever since the addition of extra
1007 controls (e.g. 4.43/32), the checks on when to allow different forms of
1008 "control" were broken. There should now be diagnostics for all cases when a
1009 control that does not make sense is encountered.
1011 7. $recipients is now available in the predata ACL (oversight).
1013 8. Tidy the search cache before the fork to do a delivery from a message
1014 received from the command line. Otherwise the child will trigger a lookup
1015 failure and thereby defer the delivery if it tries to use (for example) a
1016 cached ldap connection that the parent has called unbind on.
1018 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1019 of $address_data from the recipient verification was clobbered by the
1020 sender verification.
1022 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1023 was its contents. (It was OK if the option was not defined at all.)
1025 11. A "Completed" log line is now written for messages that are removed from
1026 the spool by the -Mrm option.
1028 12. $host_address is now set to the target address during the checking of
1029 ignore_target_hosts.
1031 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1032 being passed; this would have caused $sender_host_name to have been used if
1033 matching the list had actually called for a host name (not very likely,
1034 since this list is usually IP addresses). A host name is now passed as
1037 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1038 code that specifies a non-restarting handler (typically sigaction() in
1039 modern systems) in an attempt to fix a rare and obscure crash bug.
1041 15. Narrowed the window for a race in the daemon that could cause it to ignore
1042 SIGCHLD signals. This is not a major problem, because they are used only to
1043 wake it up if nothing else does.
1045 16. A malformed maildirsize file could cause Exim to calculate negative values
1046 for the mailbox size or file count. Odd effects could occur as a result.
1047 The maildirsize information is now recalculated if the size or filecount
1050 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1051 support for a long time. Removed HAVE_SYS_VFS_H.
1053 18. Updated exipick to current release from John Jetmore.
1055 19. Allow an empty sender to be matched against a lookup in an address list.
1056 Previously the only cases considered were a regular expression, or an
1059 20. Exim went into a mad DNS lookup loop when doing a callout where the
1060 host was specified on the transport, if the DNS lookup yielded more than
1063 21. The RFC2047 encoding function was originally intended for short strings
1064 such as real names; it was not keeping to the 75-character limit for
1065 encoded words that the RFC imposes. It now respects the limit, and
1066 generates multiple encoded words if necessary. To be on the safe side, I
1067 have increased the buffer size for the ${rfc2047: expansion operator from
1070 22. Failure to deliver a bounce message always caused it to be frozen, even if
1071 there was an errors_to setting on the router. The errors_to setting is now
1074 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1075 canonical form (fully expanded) before being placed in
1076 $sender_host_address.
1078 24. Updated eximstats to version 1.33
1080 25. Include certificate and key file names in error message when GnuTLS fails
1081 to set them up, because the GnuTLS error message doesn't include the name
1082 of the failing file when there is a problem reading it.
1084 26. Expand error message when OpenSSL has problems setting up cert/key files.
1087 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1088 (this can affect the format of dates).
1090 28. exim_tidydb, when checking for the continued existence of a message for
1091 which it has found a message-specific retry record, was not finding
1092 messages that were in split spool directories. Consequently, it was
1093 deleting retry records that should have stayed in existence.
1095 29. eximstats updated to version 1.35
1096 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1097 1.35 - bugfix such that pie charts by volume are generated correctly
1099 30. The SPA authentication driver was not abandoning authentication and moving
1100 on to the next authenticator when an expansion was forced to fail,
1101 contradicting the general specification for all authenticators. Instead it
1102 was generating a temporary error. It now behaves as specified.
1104 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1105 (the order specifies the preference for clients). The order is now AES256,
1106 AES128, 3DES, ARCFOUR128.
1108 31. Small patch to Sieve code - explicitly set From: when generating an
1111 32. Exim crashed if a remote delivery caused a very long error message to be
1112 recorded - for instance if somebody sent an entire SpamAssassin report back
1113 as a large number of 550 error lines. This bug was coincidentally fixed by
1114 increasing the size of one of Exim's internal buffers (big_buffer) that
1115 happened as part of the Exiscan merge. However, to be on the safe side, I
1116 have made the code more robust (and fixed the comments that describe what
1119 33. Some experimental protocols are using DNS PTR records for new purposes. The
1120 keys for these records are domain names, not reversed IP addresses. The
1121 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1122 leaves it alone. Component reversal etc. now happens only for IP addresses.
1125 34. The host_aton() function is supposed to be passed a string that is known
1126 to be a valid IP address. However, in the case of IPv6 addresses, it was
1127 not checking this. This is a hostage to fortune. Exim now panics and dies
1128 if the condition is not met. A case was found where this could be provoked
1129 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1130 components; fortuitously, this particular loophole had already been fixed
1131 by change 4.50/55 or 4.44/33 above.
1133 If there are any other similar loopholes, the new check in host_aton()
1134 itself should stop them being exploited. The report I received stated that
1135 data on the command line could provoke the exploit when Exim was running as
1136 exim, but did not say which command line option was involved. All I could
1137 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1138 running as the user.
1141 35. There was a buffer overflow vulnerability in the SPA authentication code
1142 (which came originally from the Samba project). I have added a test to the
1143 spa_base64_to_bits() function which I hope fixes it.
1146 36. The daemon start-up calls getloadavg() while still root for those OS that
1147 need the first call to be done as root, but it missed one case: when
1148 deliver_queue_load_max is set with deliver_drop_privilege. This is
1149 necessary for the benefit of the queue runner, because there is no re-exec
1150 when deliver_drop_privilege is set.
1152 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1153 was in use, was not putting the data itself into the right store pool;
1154 consequently, it could be overwritten for a subsequent message in the same
1155 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1158 38. Sometimes the final signoff response after QUIT could fail to get
1159 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1160 < 0 before doing a fflush(). This bug looks as though it goes back to the
1161 introduction of TLS in release 3.20, but "sometimes" must have been rare
1162 because the tests only now provoked it.
1168 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1169 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1170 Now it returns an integer. A number of calls were still expecting a T/F
1171 return. Fortuitously, in all cases, the tests worked in OK situations,
1172 which is the norm. However, things would have gone wrong on any write
1173 failures on the smtp file descriptor. This function is used when sending
1174 messages over SMTP and also when doing verify callouts.
1176 2. When Exim is called to do synchronous delivery of a locally submitted
1177 message (the -odf or -odi options), it no longer closes stderr before doing
1180 3. Implemented the mua_wrapper option.
1182 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1184 5. Implemented the functions header_remove(), header_testname(),
1185 header_add_at_position(), and receive_remove_recipient(), and exported them
1188 6. If an ACL "warn" statement specified the addition of headers, Exim already
1189 inserted X-ACL-Warn: at the start if there was no header name. However, it
1190 was not making this test for the second and subsequent header lines if
1191 there were newlines in the string. This meant that an invalid header could
1192 be inserted if Exim was badly configured.
1194 7. Allow an ACL "warn" statement to add header lines at the start or after all
1195 the Received: headers, as well as at the end.
1197 8. Added the rcpt_4xx retry error code.
1199 9. Added postmaster_mailfrom=xxx to callout verification option.
1201 10. Added mailfrom=xxxx to the callout verification option, for verify=
1204 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1205 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1207 12. Inserted some casts to stop certain compilers warning when using pointer
1208 differences as field lengths or precisions in printf-type calls (mostly
1209 affecting debugging statements).
1211 13. Added optional readline() support for -be (dynamically loaded).
1213 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1214 same clock tick as a message's arrival, so that its received time was the
1215 same as the "first fail" time on the retry record, and that message
1216 remained on the queue past the ultimate address timeout, every queue runner
1217 would try a delivery (because it was past the ultimate address timeout) but
1218 after another failure, the ultimate address timeout, which should have then
1219 bounced the address, did not kick in. This was a "< instead of <=" error;
1220 in most cases the first failure would have been in the next clock tick
1221 after the received time, and all would be well.
1223 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1224 being recognized when the domain list was tested by the match_domain
1225 condition in an expansion string.
1227 16. Added the ${str2b64: operator.
1229 17. Exim was always calling setrlimit() to set a large limit for the number of
1230 processes, without checking whether the existing limit was already
1231 adequate. (It did check for the limit on file descriptors.) Furthermore,
1232 errors from getrlimit() and setrlimit() were being ignored. Now they are
1233 logged to the main and panic logs, but Exim does carry on, to try to do its
1234 job under whatever limits there are.
1236 18. Imported PCRE 5.0.
1238 19. Trivial typo in log message " temporarily refused connection" (the leading
1241 20. If the log selector return_path_on_delivery was set and an address was
1242 redirected to /dev/null, the delivery process crashed because it assumed
1243 that a return path would always be set for a "successful" delivery. In this
1244 case, the whole delivery is bypassed as an optimization, and therefore no
1247 21. Internal re-arrangement: the function for sending a challenge and reading
1248 a response while authentication was assuming a zero-terminated challenge
1249 string. It's now changed to take a pointer and a length, to allow for
1250 binary data in such strings.
1252 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1254 23. Exim was not respecting finduser_retries when seeking the login of the
1255 uid under which it was called; it was always trying 10 times. (The default
1256 setting of finduser_retries is zero.) Also, it was sleeping after the final
1257 failure, which is pointless.
1259 24. Implemented tls_on_connect_ports.
1261 25. Implemented acl_smtp_predata.
1263 26. If the domain in control=submission is set empty, Exim assumes that the
1264 authenticated id is a complete email address when it generates From: or
1265 Sender: header lines.
1267 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1268 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1269 chown and chgrp in /bin and hostname in /usr/bin.
1271 28. Exim was keeping the "process log" file open after each use, just as it
1272 does for the main log. This opens the possibility of it remaining open for
1273 long periods when the USR1 signal hits a daemon. Occasional processlog
1274 errors were reported, that could have been caused by this. Anyway, it seems
1275 much more sensible not to leave this file open at all, so that is what now
1278 29. The long-running daemon process does not normally write to the log once it
1279 has entered its main loop, and it closes the log before doing so. This is
1280 so that log files can straightforwardly be renamed and moved. However,
1281 there are a couple of unusual error situations where the daemon does write
1282 log entries, and I had neglected to close the log afterwards.
1284 30. The text of an SMTP error response that was received during a remote
1285 delivery was being truncated at 512 bytes. This is too short for some of
1286 the long messages that one sometimes sees. I've increased the limit to
1289 31. It is now possible to make retry rules that apply only when a message has a
1290 specific sender, in particular, an empty sender.
1292 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1293 it possible to be selective about when SMTP synchronization is enforced.
1295 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1297 32. Implemented hosts_connection_nolog.
1299 33. Added an ACL for QUIT.
1301 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1304 35. Added mailbox_size and mailbox_filecount to appendfile.
1306 36. Added control = no_multiline_responses to ACLs.
1308 37. There was a bug in the logic of the code that waits for the clock to tick
1309 in the case where the clock went backwards by a substantial amount such
1310 that the microsecond fraction of "now" was more than the microsecond
1311 fraction of "then" (but the whole seconds number was less).
1313 38. Added support for the libradius Radius client library this is found on
1314 FreeBSD (previously only the radiusclient library was supported).
1320 1. When certain lookups returned multiple values in the form name=value, the
1321 quoting of the values was not always being done properly. Specifically:
1322 (a) If the value started with a double quote, but contained no whitespace,
1324 (b) If the value contained whitespace other than a space character (i.e.
1325 tabs or newlines or carriage returns) it was not quoted.
1326 This fix has been applied to the mysql and pgsql lookups by writing a
1327 separate quoting function and calling it from the lookup code. The fix
1328 should probably also be applied to nisplus, ibase and oracle lookups, but
1329 since I cannot test any of those, I have not disturbed their existing code.
1331 2. A hit in the callout cache for a specific address caused a log line with no
1332 reason for rejecting RCPT. Now it says "Previous (cached) callout
1333 verification failure".
1335 3. There was an off-by-one bug in the queryprogram router. An over-long
1336 return line was truncated at 256 instead of 255 characters, thereby
1337 overflowing its buffer with the terminating zero. As well as fixing this, I
1338 have increased the buffer size to 1024 (and made a note to document this).
1340 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1341 when Exim is waiting for an SMTP response from a remote server, Exim
1342 restarts its select() call on the socket, thereby resetting its timeout.
1343 This is not a problem when such interrupts are rare. Somebody set up a cron
1344 job to run exiwhat every 2 minutes, which is less than the normal select()
1345 timeout (5 or 10 minutes). This meant that the select() timeout never
1346 kicked in because it was always reset. I have fixed this by comparing the
1347 time when an interrupt arrives with the time at the start of the first call
1348 to select(). If more time than the timeout has elapsed, the interrupt is
1349 treated as a timeout.
1351 5. Some internal re-factoring in preparation for the addition of Sieve
1352 extensions (by MH). In particular, the "personal" test is moved to a
1353 separate function, and given an option for scanning Cc: and Bcc: (which is
1354 not set for Exim filters).
1356 6. When Exim created an email address using the login of the caller as the
1357 local part (e.g. when creating a From: or Sender: header line), it was not
1358 quoting the local part when it contained special characters such as @.
1360 7. Installed new OpenBSD configuration files.
1362 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1363 try to make them clearer.
1365 9. Callout options, other than the timeout value, were being ignored when
1366 verifying sender addresses in header lines. For example, when using
1368 verify = header_sender/callout=no_cache
1370 the cache was (incorrectly) being used.
1372 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1373 only the Cygwin environment.
1375 11. When return_path_on_delivery was set as a log selector, if different remote
1376 addresses in the same message used different return paths and parallel
1377 remote delivery occurred, the wrong values would sometimes be logged.
1378 (Whenever a remote delivery process finished, the return path value from
1379 the most recently started remote delivery process was logged.)
1381 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1382 lines when AUTH and/or TLS are in use. This is the "received protocol"
1383 field. Exim used to use "asmtp" for authenticated SMTP, without any
1384 indication (in the protocol name) for TLS use. Now it follows the RFC and
1385 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1386 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1387 names appear in log lines as well as in Received: header lines.
1389 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1390 extensions, and comparison tests, and to fix some bugs.
1392 14. Changes to the "personal" filter test:
1394 (1) The test was buggy in that it was just doing the equivalent of
1395 "contains" tests on header lines. For example, if a user's address was
1396 anne@some.where, the "personal" test would incorrectly be true for
1398 To: susanne@some.where
1400 This test is now done by extracting each address from the header in turn,
1401 and checking the entire address. Other tests that are part of "personal"
1402 are now done using regular expressions (for example, to check local parts
1403 of addresses in From: header lines).
1405 (2) The list of non-personal local parts in From: addresses has been
1406 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1407 taken from the Sieve specification recommendations.
1409 (3) If the message contains any header line starting with "List-" it is
1410 treated as non-personal.
1412 (4) The test for "circular" in the Subject: header line has been removed
1413 because it now seems ill-conceived.
1415 15. Minor typos in src/EDITME comments corrected.
1417 16. Installed latest exipick from John Jetmore.
1419 17. If headers_add on a router specified a text string that was too long for
1420 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1421 of string_sprintf() is now avoided.
1423 18. $message_body_size was not set (it was always zero) when running the DATA
1424 ACL and the local_scan() function.
1426 19. For the "mail" command in an Exim filter, no default was being set for
1427 the once_repeat time, causing a random time value to be used if "once" was
1428 specified. (If the value happened to be <= 0, no repeat happened.) The
1429 default is now 0s, meaning "never repeat". The "vacation" command was OK
1430 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1431 (I found it when inspecting the code).
1433 20. There is now an overall timeout for performing a callout verification. It
1434 defaults to 4 times the callout timeout, which applies to individual SMTP
1435 commands during the callout. The overall timeout applies when there is more
1436 than one host that can be tried. The timeout is checked before trying the
1437 next host. This prevents very long delays if there are a large number of
1438 hosts and all are timing out (e.g. when the network connections are timing
1439 out). The value of the overall timeout can be changed by specifying an
1440 additional sub-option for "callout", called "maxwait". For example:
1442 verify = sender/callout=5s,maxwait=20s
1444 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1445 to the end before writing, but this should make it even safer).
1447 22. Exim was forgetting that it had advertised PIPELINING for the second and
1448 subsequent messages on an SMTP connection. It was also not resetting its
1449 memory on STARTTLS and an internal HELO.
1451 23. When Exim logs an SMTP synchronization error within a session, it now
1452 records whether PIPELINING has been advertised or not.
1454 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1455 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1456 rather than long int.
1458 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1460 26. Added the never_mail option to autoreply.
1466 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1467 crash if the getsockname() call failed; this can happen if a connection is
1468 closed very soon after it is established. The problem was simply in the
1469 order in which certain operations were done, causing Exim to try to write
1470 to the SMTP stream before it had set up the file descriptor. The bug has
1471 been fixed by making things happen in the correct order.
1477 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1478 before the connection was closed, thus losing the rejection response.
1480 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1481 some early Solaris releases, but causes trouble in current releases where
1482 socklen_t is defined.
1484 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1487 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1490 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1491 that releasing the top of it at the end releases what was used for sub-
1492 expansions (unless the block got too big). However, discard this block if
1493 the first thing is a variable or header, so that we can use its block when
1494 it is dynamic (useful for very large $message_headers, for example).
1496 6. Lookups now cache *every* query, not just the most recent. A new, separate
1497 store pool is used for this. It can be recovered when all lookup caches are
1498 flushed. Lookups now release memory at the end of their result strings.
1499 This has involved some general refactoring of the lookup sources.
1501 7. Some code has been added to the store_xxx() functions to reduce the amount
1502 of flapping under certain conditions.
1504 8. log_incoming_interface used to affect only the <= reception log lines. Now
1505 it causes the local interface and port to be added to several more SMTP log
1506 lines, for example "SMTP connection from", and rejection lines.
1508 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1510 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1512 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1513 could be overwritten at the end of the current message (or the start of a
1514 new message if it was set in a HELO ACL). The value is now preserved for
1515 the duration of the SMTP connection.
1517 12. If a transport had a headers_rewrite setting, and a matching header line
1518 contained an unqualified address, that address was qualified, even if it
1519 did not match any rewriting rules. The underlying bug was that the values
1520 of the flags that permit the existence of unqualified sender and recipient
1521 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1522 non-local messages, and by -bnq for local messages) were not being
1523 preserved with the message after it was received.
1525 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1526 "next input=" as part of the text comprising the host identity instead of
1527 the correct text. The code was using the same buffer for two different
1528 strings. However, depending on which order the printing function evaluated
1529 its arguments, the bug did not always show up. Under Linux, for example, my
1530 test suite worked just fine.
1532 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1533 doesn't work with some older versions of Perl. It has been changed to "my",
1534 which in any case is probably the better facility to use.
1536 15. A really picky compiler found some instances of statements for creating
1537 error messages that either had too many or two few arguments for the format
1540 16. The size of the buffer for calls to the DNS resolver has been increased
1541 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1542 for addresses that have a lot of PTR records. This alleviates a problem; it
1543 does not fully solve it.
1545 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1546 buffer now truncates the list and logs the incident, which is the same
1547 action as happens when Exim is looking up a host name and its aliases.
1548 Previously in this situation something unpredictable would happen;
1549 sometimes it was "internal error: store_reset failed".
1551 18. If a server dropped the connection unexpectedly when an Exim client was
1552 using GnuTLS and trying to read a response, the client delivery process
1553 crashed while trying to generate an error log message.
1555 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1556 string, for example:
1558 warn message = H1: something\nH2: something
1560 the text was added as a single header line from Exim's point of view
1561 though it ended up OK in the delivered message. However, searching for the
1562 second and subsequent header lines using $h_h2: did not work. This has been
1563 fixed. Similarly, if a system filter added multiple headers in this way,
1564 the routers could not see them.
1566 20. Expanded the error message when iplsearch is called with an invalid key to
1567 suggest using net-iplsearch in a host list.
1569 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1570 ACLs are no longer actually imposed (and a message to that effect is
1573 22. If a "gecos" field in a passwd entry contained escaped characters, in
1574 particular, if it contained a \" sequence, Exim got it wrong when building
1575 a From: or a Sender: header from that name. A second bug also caused
1576 incorrect handling when an unquoted " was present following a character
1577 that needed quoting.
1579 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1580 was not being matched caselessly.
1582 24. Arranged for all hyphens in the exim.8 source to be escaped with
1585 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1586 properly. Recipient callout cache records were still being keyed to include
1587 the sender, even when use_sender was set false. This led to far more
1588 callouts that were necessary. The sender is no longer included in the key
1589 when use_sender is false.
1591 26. Added "control = submission" modifier to ACLs.
1593 27. Added the ${base62d: operator to decode base 62 numbers.
1595 28. dnsdb lookups can now access SRV records.
1597 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1598 the configuration file.
1600 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1601 (-v) mode. This makes the output for a verbose queue run more intelligible.
1603 31. Added a use_postmaster feature to recipient callouts.
1605 32. Added the $body_zerocount variable, containing the number of binary zero
1606 bytes in the message body.
1608 33. The time of last modification of the "new" subdirectory is now used as the
1609 "mailbox time last read" when there is a quota error for a maildir
1612 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1614 35. Added +ignore_unknown as a special item in host lists.
1616 36. Code for decoding IPv6 addresses in host lists is now included, even if
1617 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1618 address was recognized as an IP address, but was then not correctly decoded
1619 into binary, causing unexpected and incorrect effects when compared with
1626 1. Very minor rewording of debugging text in manualroute to say "list of
1627 hosts" instead of "hostlist".
1629 2. If verify=header_syntax was set, and a header line with an unqualified
1630 address (no domain) and a large number of spaces between the end of the
1631 name and the colon was received, the reception process suffered a buffer
1632 overflow, and (when I tested it) crashed. This was caused by some obsolete
1633 code that should have been removed. The fix is to remove it!
1635 3. When running in the test harness, delay a bit after writing a bounce
1636 message to get a bit more predictability in the log output.
1638 4. Added a call to search_tidyup() just before forking a reception process. In
1639 theory, someone could use a lookup in the expansion of smtp_accept_max_
1640 per_host which, without the tidyup, could leave open a database connection.
1642 5. Added the variables $recipient_data and $sender_data which get set from a
1643 lookup success in an ACL "recipients" or "senders" condition, or a router
1644 "senders" option, similar to $domain_data and $local_part_data.
1646 6. Moved the writing of debug_print from before to after the "senders" test
1649 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1650 problems for message scanning, either using a data ACL, or using
1651 local_scan() because the Received: header was not generated till after they
1652 were called (in order to set the time as the time of reception completion).
1653 I have revised the way this works. The header is now generated after the
1654 body is received, but before the ACL or local_scan() are called. After they
1655 are run, the timestamp in the header is updated.
1661 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1662 before starting a queue runner without re-exec. This happened only when
1663 deliver_drop_privilege was set or when the Exim user was set to root. The
1664 effect of the bug was that timeouts during subsequent deliveries caused
1665 crashes instead of being properly handled. The handler is now left at its
1666 default (and expected) setting.
1668 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1669 message, again when deliver_drop_privilege is set or Exim is run as root.
1670 The bug described in (1) was not present in this case, but the tidying up
1671 of the other signals was missing. I have made the two cases consistent.
1673 3. The ignore_target_hosts setting on a manualroute router was being ignored
1674 for hosts that were looked up using the /MX notation.
1676 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1679 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1680 operated on the sender address. After changing the $sender_address to <>
1681 for the sender address verify, Exim was re-instated it as the original
1682 (before rewriting) address, but remembering that it had rewritten it, so it
1683 wasn't rewriting it again. This bug also had the effect of breaking the
1684 sender address verification caching when the sender address was rewritten.
1686 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1687 This has been changed so that if the ip literal address matches
1688 ignore_target_hosts, the router declines.
1690 7. Added expansion conditions match_domain, match_address, and match_local_
1691 part (NOT match_host).
1693 8. The placeholder for the Received: header didn't have a length field set.
1695 9. Added code to Exim itself and to exim_lock to test for a specific race
1696 condition that could lead to file corruption when using MBX delivery. The
1697 issue is with the lockfile that is created in /tmp. If this file is removed
1698 after a process has opened it but before that process has acquired a lock,
1699 there is the potential for a second process to recreate the file and also
1700 acquire a lock. This could lead to two Exim processes writing to the file
1701 at the same time. The added code performs the same test as UW imapd; it
1702 checks after acquiring the lock that its file descriptor still refers to
1703 the same named file.
1705 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1706 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1709 11. Added the smtp_active_hostname option. If used, this will typically be made
1710 to depend on the incoming interface address. Because $interface_address is
1711 not set up until the daemon has forked a reception process, error responses
1712 that can happen earlier (such as "too many connections") no longer contain
1715 12. If an expansion in a condition on a "warn" statement fails because a lookup
1716 defers, the "warn" statement is abandoned, and the next ACL statement is
1717 processed. Previously this caused the whole ACL to be aborted.
1719 13. Added the iplsearch lookup type.
1721 14. Added ident_timeout as a log selector.
1723 15. Added tls_certificate_verified as a log selector.
1725 16. Added a global option tls_require_ciphers (compare the smtp transport
1726 option of the same name). This controls incoming TLS connections.
1728 17. I finally figured out how to make tls_require_ciphers do a similar thing
1729 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1730 before starting the TLS session.
1732 18. Tabs are now shown as \t in -bP output.
1734 19. If the log selector return_path_on_delivery was set, Exim crashed when
1735 bouncing a message because it had too many Received: header lines.
1737 20. If two routers both had headers_remove settings, and the first one included
1738 a superfluous trailing colon, the final name in the first list and the
1739 first name in the second list were incorrectly joined into one item (with a
1740 colon in the middle).
1746 1. Added -C and -D options to the exinext utility, mainly to make it easier
1747 to include in the automated testing, but these could be helpful when
1748 multiple configurations are in use.
1750 2. The exinext utility was not formatting the output nicely when there was
1751 an alternate port involved in the retry record key, nor when there was a
1752 message id as well (for retries that were specific to a specific message
1753 and a specific host). It was also confused by IPv6 addresses, because of
1754 the additional colons they contain. I have fixed the IPv4 problem, and
1755 patched it up to do a reasonable job for IPv6.
1757 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1758 delivery, the log line now contains "pipelined" if PIPELINING was used.
1760 4. An SMTP transport process used to panic and die if the bind() call to set
1761 an explicit outgoing interface failed. This has been changed; it is now
1762 treated in the same way as a connect() failure.
1764 5. A reference to $sender_host_name in the part of a conditional expansion
1765 that was being skipped was still causing a DNS lookup. This no longer
1768 6. The def: expansion condition was not recognizing references to header lines
1769 that used bh_ and bheader_.
1771 7. Added the _cache feature to named lists.
1773 8. The code for checking quota_filecount in the appendfile transport was
1774 allowing one more file than it should have been.
1776 9. For compatibility with Sendmail, the command line option
1784 and sets the incoming protocol and host name (for trusted callers). The
1785 host name and its colon can be omitted when only the protocol is to be set.
1786 Note the Exim already has two private options, -pd and -ps, that refer to
1787 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1788 "s", but I don't think that's a major issue.
1790 10. A number of refactoring changes to the code, none of which should affect
1793 (a) The number of logging options was getting close to filling up the
1794 32-bit word that was used as a bit map. I have split them into two classes:
1795 those that are passed in the argument to log_write(), and those that are
1796 only ever tested independently outside of that function. These are now in
1797 separate 32-bit words, so there is plenty of room for expansion again.
1798 There is no change in the user interface or the logging behaviour.
1800 (b) When building, for example, log lines, the code previously used a
1801 macro that called string_cat() twice, in order to add two strings. This is
1802 not really sufficiently general. Furthermore, there was one instance where
1803 it was actually wrong because one of the argument was used twice, and in
1804 one call a function was used. (As it happened, calling the function twice
1805 did not affect the overall behaviour.) The macro has been replaced by a
1806 function that can join an arbitrary number of extra strings onto a growing
1809 (c) The code for expansion conditions now uses a table and a binary chop
1810 instead of a serial search (which was left over from when there were very
1811 few conditions). Also, it now recognizes conditions like "pam" even when
1812 the relevant support is not compiled in: a suitably worded error message is
1813 given if an attempt is made to use such a condition.
1815 11. Added ${time_interval:xxxxx}.
1817 12. A bug was causing one of the ddress fields not to be passed back correctly
1818 from remote delivery subprocesses. The field in question was not being
1819 subsequently used, so this caused to problems in practice.
1821 13. Added new log selectors queue_time and deliver_time.
1823 14. Might have fixed a bug in maildirsizefile handling that threw up
1824 "unexpected character" debug warnings, and recalculated the data
1825 unnecessarily. In any case, I expanded the warning message to give more
1828 15. Added the message "Restricted characters in address" to the statements in
1829 the default ACL that block characters like @ and % in local parts.
1831 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1832 Three changes have been made:
1834 (a) There was a serious bug; a negative response to MAIL caused the whole
1835 recipient domain to be cached as invalid, thereby blocking all messages
1836 to all local parts at the same domain, from all senders. This bug has
1837 been fixed. The domain is no longer cached after a negative response to
1838 MAIL if the sender used is not empty.
1840 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1843 (c) A new callout option, "use_sender" has been added for people who want
1844 the modified behaviour.
1850 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1853 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1854 indeed breaks things for older releases.
1856 3. Added additional logging to the case where there is a problem reading data
1857 from a filter that is running in a subprocess using a pipe, in order to
1858 try to track down a specific problem.
1860 4. Testing facility fudge: when running in the test harness and attempting
1861 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1862 getting "No route to host". Convert this to a timeout.
1864 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1867 6. Some OS don't have socklen_t but use size_t instead. This affects the
1868 fifth argument of getsockopt() amongst other things. This is now
1869 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1870 can be set for individual OS. I have set it for SunOS5, OSF1, and
1871 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1872 some earlier ones do not.
1874 7. Change 4.30/15 was not doing the test caselessly.
1876 8. The standard form for an IPv6 address literal was being rejected by address
1877 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1878 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1879 this, as well as the form without the "IPv6" on the front (but only when
1880 address literals are enabled, of course).
1882 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1884 10. Exim crashed if a message with an empty sender address specified by -f
1885 encountered a router with an errors_to setting. This could be provoked only
1886 by a command such as
1890 where an empty string was supplied; "<>" did not hit this bug.
1892 11. Installed PCRE release 4.5.
1894 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1895 remained set. It is now erased.
1897 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1898 times from message ids (which are base 36 rather than the normal 62).
1900 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1901 were being counted as actual protocol errors, and logged if the log
1902 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1903 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1904 and DATA following a set of rejected RCPTs do not count as protocol errors.
1905 In other words, Exim assumes they were pipelined, though this may not
1906 actually be the case. Of course, in all cases the client gets an
1907 appropriate error code.
1909 15. If a lookup fails in an ACL condition, a message about the failure may
1910 be available; it is used if testing the ACL cannot continue, because most
1911 such messages specify what the cause of the deferral is. However, some
1912 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1913 that caused an old message to be retained and used if a later statement
1914 caused a defer, replacing the real cause of the deferral.
1916 16. If an IP address had so many PTR records that the DNS lookup buffer
1917 was not large enough to hold them, Exim could crash while trying to process
1918 the truncated data. It now detects and logs this case.
1920 17. Further to 4.21/58, another change has been made: if (and only if) the
1921 first line of a message (the first header line) ends with CRLF, a bare LF
1922 in a subsequent header line has a space inserted after it, so as not to
1923 terminate the header.
1925 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1926 unnecessarily, used atoi() instead of strtol(), and didn't check the
1927 termination when getting file sizes from file names by regex.
1929 19. Completely re-implemented the support for maildirsize files, in the light
1930 of a number of problems with the previous contributed implementation
1931 (4.30/29). In particular:
1933 . If the quota is zero, the maildirsize file is maintained, but no quota is
1936 . If the maildir directory does not exist, it is created before any attempt
1937 to write a maildirsize file.
1939 . The quota value in the file is just a cache; if the quota is changed in
1940 the transport, the new value overrides.
1942 . A regular expression is available for excluding directories from the
1945 20. The autoreply transport checks the characters in options that define the
1946 message's headers; it allows continued headers, but it was checking with
1947 isspace() after an embedded newline instead of explicitly looking for a
1950 21. If all the "regular" hosts to which an address was routed had passed their
1951 expiry times, and had not reached their retry times, the address was
1952 bounced, even if fallback hosts were defined. Now Exim should go on to try
1955 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1956 equivalent code in the SMTP transport. Some hosts send humungous responses
1957 to HELO/EHLO, more than 1024 it seems.
1959 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1960 gives compiler warnings in some environments. I've now done it "properly",
1963 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1964 (because of problems with the built-in one) was declared to return uschar *
1965 instead of char *, causing compiler failure.
1967 25. Fixed a file descriptor leak when processing alias/forward files.
1969 26. Fixed a minor format string issue in dbfn.c.
1971 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1973 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1974 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1977 29. When a sender address is verified, it is cached, to save repeating the test
1978 when there is more than one recipient in a message. However, when the
1979 verification involves a callout, it is possible for different callout
1980 options to be set for different recipients. It is too complicated to keep
1981 track of this in the cache, so now Exim always runs a verification when a
1982 callout is required, relying on the callout cache for the optimization.
1983 The overhead is duplication of the address routing, but this should not be
1986 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
1987 to be verified with callout=postmaster, and the main callout worked but the
1988 postmaster check failed, the verification correctly failed. However, if a
1989 subsequent RCPT command asked for sender verification *without* the
1990 postmaster check, incorrect caching caused this verification also to fail,
1993 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
1994 it was not caching the DNS options (qualify_single, search_parents) that
1995 were used when the lookup failed. A subsequent lookup with different
1996 options therefore always gave the same answer, though there were cases
1997 where it should not have. (Example: a "domains = !$mx_any" option on a
1998 dnslookup router: the "domains" option is always processed without any
1999 widening, but the router might have qualify_single set.) Now Exim uses the
2000 cached value only when the same options are set.
2002 32. Added John Jetmore's "exipick" utility to the distribution.
2004 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2005 than a timeout (e.g. a certificate is required, and is not provided), an
2006 Exim server now closes the connection immediately. Previously it waited for
2007 the client to close - but if the client is SSL, it seems that they each
2008 wait for each other, leading to a delay before one of them times out.
2010 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2011 maintained 0.8.x compatibility because I don't think many are using it, and
2012 it is clearly obsolete.
2014 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2017 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2018 client certificate was expired. A simple patch fixes this, though I don't
2019 understand the full logic of why the verify callback is called multiple
2022 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2023 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2024 which causes problems with some clients (such as the Certicom SSL Plus
2025 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2026 disables the coutermeasure allowing Eudora to connect."
2028 38. Exim was not checking that a write() to a log file succeeded. This could
2029 lead to Bad Things if a log got too big, in particular if it hit a file
2030 size limit. Exim now panics and dies if it cannot write to a log file, just
2031 as it does if it cannot open a log file.
2033 39. Modified OS/Makefile-Linux so that it now contains
2035 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2037 The two -D definitions ensure that Exim is compiled with large file
2038 support, which makes it possible to handle log files that are bigger than
2041 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2042 instance) a domain was checked against a named list that involved a lookup,
2043 causing $domain_data to be set, then another domain was checked against the
2044 same list, then the first domain was re-checked, the value of $domain_data
2045 after the final check could be wrong. In particular, if the second check
2046 failed, it could be set empty. This bug probably also applied to
2049 41. The strip_trailing_dot option was not being applied to the address given
2050 with the -f command-line option.
2052 42. The code for reading a message's header from the spool was incrementing
2053 $received_count, but never initializing it. This meant that the value was
2054 incorrect (doubled) while delivering a message in the same process in which
2055 it was received. In the most common configuration of Exim, this never
2056 happens - a fresh exec is done - but it can happen when
2057 deliver_drop_privilege is set.
2059 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2060 it now includes up to 150 characters of the unexpected data in the log
2063 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2064 and building data strings. The size of both of these buffers was 10 000
2065 bytes - far larger than anybody would *ever* want, thought I. Needless to
2066 say, somebody hit the limit. I have increased the maximum line length to
2067 20 000 and the maximum data length of concatenated lines to 100 000. I have
2068 also fixed two bugs, because there was no checking on these buffers. Tsk,
2069 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2070 buffer is too small.
2072 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2073 lsearch lookups. Now it does.
2075 46. When parsing a route_list item in a manualroute router, a fixed-length
2076 buffer was used for the list of hosts. I made this 1024 bytes long,
2077 thinking that nobody would ever have a list of hosts that long. Wrong.
2078 Somebody had a whole pile of complicated expansion conditions, and the
2079 string was silently truncated, leading to an expansion error. It turns out
2080 that it is easier to change to an unlimited length (owing to other changes
2081 that have happened since this code was originally written) than to build
2082 structure for giving a limitation error. The length of the item that
2083 expands into the list of hosts is now unlimited.
2085 47. The lsearch lookup could not handle data where the length of text line was
2086 more than 4095 characters. Such lines were truncated, leading to shortened
2087 data being returned. It should now handle lines of any length.
2089 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2090 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2093 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2094 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2095 no longer happens in comment lines. A list of replacements is now placed
2096 at the head of all of the source files, except those whose only change is
2097 to replace PERL_COMMAND in the very first #! line.
2099 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2100 messages on the queue, with a bottom-up merge sort, using code contributed
2101 by Michael Haardt. This should make operations like -bp somewhat faster on
2102 large queues. It won't affect queue runners, except when queue_run_in_order
2105 51. Installed eximstats 1.31 in the distribution.
2107 52. Added support for SRV lookups to the dnslookup router.
2109 53. If an ACL referred to $message_body or $message_body_end, the value was not
2110 reset for any messages that followed in the same SMTP session.
2112 54. The store-handling optimization for building very long strings was not
2113 differentiating between the different store pools. I don't think this
2114 actually made any difference in practice, but I've tidied it.
2116 55. While running the routers to verify a sender address, $sender_address
2117 was still set to the sender address. This is wrong, because when routing to
2118 send a bounce to the sender, it would be empty. Therefore, I have changed
2119 it so that, while verifying a sender address, $sender_address is set to <>.
2120 (There is no change to what happens when verifying a recipient address.)
2122 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2123 target A or AAAA records (if not already returned) without resetting the
2124 qualify_single or search_parents options of the DNS resolver. These are
2125 inappropriate in this case because the targets of MX and SRV records must
2126 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2127 target that, when qualified, matched something in the local domain. These
2128 two options are now turned off when doing these lookups.
2130 57. It seems that at least some releases of Reiserfs (which does not have the
2131 concept of a fixed number of inodes) returns zero and not -1 for the
2132 number of available inodes. This interacted badly with check_spool_inodes,
2133 which assumed that -1 was the "no such thing" setting. What I have done is
2134 to check that the total number of inodes is greater than zero before doing
2135 the test of how many are available.
2137 58. When a "warn" ACL statement has a log_message modifier, the message is
2138 remembered, and not repeated. This is to avoid a lot of repetition when a
2139 message has many recipients that cause the same warning to be written.
2140 Howewer, Exim was preserving the list of already written lines for an
2141 entire SMTP session, which doesn't seem right. The memory is now reset if a
2142 new message is started.
2144 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2145 debugging output unless log_rewrite was also set.
2147 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2148 of (int)(handle) when we know that handle contains (void *)(-1).
2150 61. The Exim daemon panic-logs an error return when it closes the incoming
2151 connection. However "connection reset by peer" seems to be common, and
2152 isn't really an error worthy of noting specially, so that particular error
2155 62. When Exim is trying to find all the local interfaces, it used to panic and
2156 die if the ioctl to get the interface flags failed. However, it seems that
2157 on at least one OS (Solaris 9) it is possible to have an interface that is
2158 included in the list of interfaces, but for which you get a failure error
2159 for this call. This happens when the interface is not "plumbed" into a
2160 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2161 failure of the "get flags" call assumes that the interface is down.
2163 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2164 makes life easier for people who are doing arithmetic on fields extracted
2165 from dates, where you often get leading zeros that should not be
2166 interpreted as octal.
2168 64. Added qualify_domain to the redirect router, to override the global
2171 65. If a pathologically long header line contained very many addresses (the
2172 report of this problem mentioned 10 000) and each of them was rewritten,
2173 Exim could use up a very large amount of memory. (It kept on making new
2174 copies of the header line as it rewrote, and never released the old ones.)
2175 At the expense of a bit more processing, the header rewriting function has
2176 been changed so that it no longer eats memory in this way.
2178 66. The generation of the Received: header has been moved from the time that a
2179 message starts to be received, to the time that it finishes. The timestamp
2180 in the Received: header should now be very close to that of the <= log
2181 line. There are two side-effects of this change:
2183 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2184 logged header lines no longer include the local Received: line, because
2185 it has not yet been created. The same applies to a copy of the message
2186 that is returned to a non-SMTP sender when a message is rejected.
2188 (b) When a filter file is tested using -bf, no additional Received: header
2189 is added to the test message. After some thought, I decided that this
2192 This change does not affect the value of $received_for. It is still set
2193 after address rewriting, but before local_scan() is called.
2195 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2197 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2198 gave an unhelpful panic error message, and a defer error. I have managed to
2199 change this behaviour so that it now rejects any supplied certificate,
2200 which seems right, as the list of acceptable certificates is empty.
2202 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2203 gave an unhelpful defer error. I have not managed to make this reject any
2204 supplied certificates, but the error message it gives is "no certificate
2205 supplied", which is not helpful.
2207 70. exigrep's output now also includes lines that are not associated with any
2208 message, but which match the given pattern. Implemented by a patch from
2209 Martin Sluka, which also tidied up the Perl a bit.
2211 71. Recipient callout verification, like sender verification, was using <> in
2212 the MAIL FROM command. This isn't really the right thing, since the actual
2213 sender may affect whether the remote host accepts the recipient or not. I
2214 have changed it to use the actual sender in the callout; this means that
2215 the cache record is now keyed on a recipient/sender pair, not just the
2216 recipient address. There doesn't seem to be a real danger of callout loops,
2217 since a callout by the remote host to check the sender would use <>.
2218 [SEE ABOVE: changed after hitting problems.]
2220 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2221 temporary errors. However, in the case of such a code being given after
2222 the end of a data transmission (i.e. after ".") Exim was failing to write
2223 a retry record for the message. (Yes, there was some broken host that was
2224 actually sending 8xx at this point.)
2226 73. An unknown lookup type in a host list could cause Exim to panic-die when
2227 the list was checked. (An example that provoked this was putting <; in the
2228 middle of a list instead of at the start.) If this happened during a DATA
2229 ACL check, a -D file could be left lying around. This kind of configuration
2230 error no longer causes Exim to die; instead it causes a defer errror. The
2231 incident is still logged to the main and panic logs.
2233 74. Buglet left over from Exim 3 conversion. The message "too many messages
2234 in one connection" was written to the rejectlog but not the mainlog, except
2235 when address rewriting (yes!) was being logged.
2237 75. Added write_rejectlog option.
2239 76. When a system filter was run not as root (that is, when system_filter_user
2240 was set), the values of the $n variables were not being returned to the
2241 main process; thus, they were not subsequently available in the $sn
2244 77. Added +return_path_on_delivery log selector.
2246 78. A connection timeout was being treated differently from recipients deferred
2247 when testing hosts_max_try with a message that was older than the host's
2248 retry timeout. (The host should not be counted, thus allowing all hosts to
2249 be tried at least once before bouncing.) This may have been the cause of an
2250 occasionally reported bug whereby a message would remain on the queue
2251 longer than the retry timeout, but would be bounced if a delivery was
2252 forced. I say "may" because I never totally pinned down the problem;
2253 setting up timeout/retry tests is difficult. See also the next item.
2255 79. The ultimate address timeout was not being applied to errors that involved
2256 a combination of host plus message (for example, a timeout on a MAIL
2257 command). When an address resolved to a number of possible hosts, and they
2258 were not all tried for each delivery (e.g. because of hosts_max_try), a
2259 message could remain on the queue longer than the retry timeout.
2261 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2264 81. Fixed an obscure SMTP outgoing bug which required at least the following
2265 conditions: (a) there was another message waiting for the same server;
2266 (b) the server returned 5xx to all RCPT commands in the first message so
2267 that the message was not completed; (c) the server dropped the connection
2268 or gave a negative response to the RSET that Exim sends to abort the
2269 transaction. The observed case was a dropped connection after DATA that had
2270 been sent in pipelining mode. That is, the server had advertised PIPELINING
2271 but was not implementing it correctly. The effect of the bug was incorrect
2272 behaviour, such as trying another host, and this could lead to a crash.
2278 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2279 and daemon.c were passed as pointers to ints; they should have been
2280 pointers to socklen_t variables (which are typically unsigned ints).
2282 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2285 3. Fixed a really odd bug that affected only the testing scheme; patching a
2286 certain fixed string in the binary changed the value of another string that
2287 happened to be identical to the end of the original first string.
2289 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2290 name", it returns that address as the IP address. On some operating
2291 systems (e.g. Solaris), it also passes back the IP address string as the
2292 "host name". However, on others (e.g. Linux), it passes back an empty
2293 string. Exim wasn't checking for this, and was changing the host name to an
2294 empty string, assuming it had been canonicized.
2296 5. Although rare, it is permitted to have more than one PTR record for a given
2297 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2298 all the names associated with an address, because they do in Solaris.
2299 However, it seems that they do not in Linux for data that comes from the
2300 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2301 I found this out when I moved to a new Linux workstation and tried to run
2302 the Exim test suite.
2304 To get round this problem I have changed the code so that it now does its
2305 own call to the DNS to look up PTR records when searching for a host name.
2306 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2307 addresses that are only in /etc/hosts are still found.
2309 This behaviour is, however, controlled by an option called host_lookup_
2310 order, which defaults to "bydns:byaddr". If people want to use the other
2311 order, or indeed, just use one or the other means of lookup, they can
2312 specify it in this variable.
2314 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2315 some operating systems, this comes back from gethostbyaddr() as an empty
2316 string, and this is what Exim used to test for. However, it seems that in
2317 other systems, "." is yielded. Exim now tests for this case too.
2319 7. The values of check_spool_space and check_log_space are now held internally
2320 as a number of kilobytes instead of an absolute number of bytes. If a
2321 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2322 kilobyte. This means that much larger values can be stored.
2324 8. Exim monitor: an attempt to get the action menu when not actually pointing
2325 at a message produces an empty menu entitled "No message selected". This
2326 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2327 no entries in it ("Shell widget menu has zero width and/or height"). So I
2328 have added a single, blank menu entry in this case.
2330 9. Added ${quote_local_part.
2332 10. MIME decoding is now applied to the contents of Subject: header lines when
2335 11. Now that a reference to $sender_host_address automatically causes a reverse
2336 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2337 host lookup before query-style lookups in lists that might use this
2338 variable. This has therefore been abolished, and the "net-" prefix is no
2339 longer necessary for query-style lookups.
2341 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2342 have been a typo for LFLAGS, so it has been changed.
2344 13. The install script calls Exim with "-C /dev/null" in order to find the
2345 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2346 to be output. Howeve, since Exim outputs its version number before the
2347 error, it didn't break the script. It just looked ugly. I fixed this by
2348 always allowing "-C /dev/null" if the caller is root.
2350 14. Ignore overlarge ACL variable number when reading spool file - insurance
2351 against a later release with more variables having written the file.
2353 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2354 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2355 this, as well as the form without the "IPv6" on the front.
2357 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2358 OS/Makefile-Darwin file.
2360 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2361 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2363 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2364 variable. However, some OS use ints for inodes. I've added cast to long int
2365 to get rid of the compiler warning.
2367 19. I had forgotten to lock out "/../" in configuration file names when
2368 ALT_CONFIG_PREFIX was set.
2370 20. Routers used for verification do not need to specify transports. However,
2371 if such a router generated a host list, and callout was configured, Exim
2372 crashed, because it could not find a port number from the (non-existent)
2373 transport. It now assumes port 25 in this circumstance.
2375 21. Added the -t option to exigrep.
2377 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2378 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2379 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2380 them is set, LOOKUP_LSEARCH is forced.
2382 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2384 24. Added sender and host information to the "rejected by local_scan()" log
2385 line; previously there was no indication of these.
2387 25. Added .include_if_exists.
2389 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2390 Linux. It turns out that not all file systems support this. Apparently some
2391 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2392 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2393 not supported on the file descriptor, is now ignored when Exim is trying to
2394 sync a directory. This applies only to Linux.
2396 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2398 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2399 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2402 29. Added support for maildirsize files from supplied patch (modified a bit).
2404 30. The use of :fail: followed by an empty string could lead Exim to respond to
2405 sender verification failures with (e.g.):
2407 550 Verification failed for <xxx>
2408 550 Sender verify failed
2410 where the first response line was missing the '-' that indicates it is not
2411 the final line of the response.
2413 31. The loop for finding the name of the user that called Exim had a hardwired
2414 limit of 10; it now uses the value of finduser_retries, which is used for
2415 all other user lookups.
2417 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2420 33. Exim was neglecting to zero errno before one call of strtol() when
2421 expanding a string and expecting an integer value. On some systems this
2422 resulted in spurious "integer overflow" errors. Also, it was casting the
2423 result into an int without checking.
2425 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2426 did not work. The code looks as if it has *never* worked, though it appears
2427 to have been documented since at least releast 1.62. I have made it work.
2429 35. The "timeout_DNS" error in retry rules, also documented since at least
2430 1.62, also never worked. As it isn't clear exactly what this means, and
2431 clearly it isn't a major issue, I have abolished the feature by treating it
2432 as "timeout", and writing a warning to the main and panic logs.
2434 36. The display of retry rules for -brt wasn't always showing the error code
2437 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2438 timeout_connect_A, timeout_connect_MX.
2440 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2441 to the empty sender.
2443 39. The daemon was not analysing the content of -oX till after it had closed
2444 stderr and disconnected from the controlling terminal. This meant that any
2445 syntax errors were only noted on the panic log, and the return code from
2446 the command was 0. By re-arranging the code a little, I've made the
2447 decoding happen first, so such errors now appear on stderr, and the return
2448 code is 1. However, the actual setting up of the sockets still happens in
2449 the disconnected process, so errors there are still only recorded on the
2452 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2453 connections (as happens on some IP stacks) was logged at start up time as
2454 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2455 it from "IPv6 and IPv4", which means that two separate sockets are being
2458 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2459 says whether AF_INET or AF_INET6 was passed as an argument.
2461 42. Exiwhat output was messed up when time zones were included in log
2464 43. Exiwhat now gives more information about the daemon's listening ports,
2465 and whether -tls-on-connect was used.
2467 44. The "port" option of the smtp transport is now expanded.
2469 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2470 silently ignored. Now an error message is written to the main and panic
2473 46. There's a new ACL modifier called "logwrite" which writes to a log file
2474 as soon as it is encountered.
2476 47. Added $local_user_uid and $local_user_gid at routing time.
2478 48. Exim crashed when trying to verify a sender address that was being
2481 49. Exim was recognizing only a space character after ".include". It now also
2482 recognizes a tab character.
2484 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2485 extracting the relevant information from the specification. The man page no
2486 longer contains scrambled data for the -d option, and I've added a section
2487 at the front about calling Exim under different names.
2489 51. Added "extra_headers" argument to the "mail" command in filter files.
2491 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2494 53. Installed eximstats 1.29.
2496 54. Added transport_filter_timeout as a generic transport option.
2498 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2499 Cc: header lines. This was required by RFC 822, but it not required by RFC
2502 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2503 incoming messages that did not have them. Now it does so only if the
2504 message originates locally, that is, if there is no associated remote host
2505 address. When Resent- header lines are present, this applies to the Resent-
2506 lines rather than the non-Resent- lines.
2508 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2509 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2511 58. Messages for configuration errors now include the name of the main
2512 configuration file - useful now that there may be more than one file in a
2513 list (.included file names were always shown).
2515 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2516 for those rare installations that do not start the daemon as root or run it
2517 setuid root. I've cut out the call to initgroups() if the daemon is not
2520 60. The Exim user and group can now be bound into the binary as text strings
2521 that are looked up at the start of Exim's processing.
2523 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2525 62. Added $mailstore_basename variable.
2527 63. Installed patch to sieve.c from Michael Haardt.
2529 64. When Exim failed to open the panic log after failing to open the main log,
2530 the original message it was trying to log was written to stderr and debug
2531 output, but if they were not available (the usual case in production), it
2532 was lost. Now it is written to syslog before the two lines that record the
2533 failures to open the logs.
2535 65. Users' Exim filters run in subprocesses under the user's uid. It is
2536 possible for a "deliver" command or an alias in a "personal" command to
2537 provoke an address rewrite. If logging of address rewriting is configured,
2538 this fails because the process is not running as root or exim. There may be
2539 a better way of dealing with this, but for the moment (because 4.30 needs
2540 to be released), I have disabled address rewrite logging when running a
2541 filter in a non-root, non-exim process.
2547 1. The buildconfig auxiliary program wasn't quoting the value set for
2548 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2549 not defined. This bug was masked in 4.22 by the effect that was fixed in
2552 2. Some messages that were rejected after a message id was allocated were
2553 shown as "incomplete" by exigrep. It no longer does this for messages that
2554 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2556 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2557 have allow_domain_literals set, the ID did not get logged in the <= line.
2558 Domain literals are now always recognized in Message-ID: header lines.
2560 4. The first argument for a ${extract expansion item is the key name or field
2561 number. Leading and trailing spaces in this item were not being ignored,
2562 causing some misleading effects.
2564 5. When deliver_drop_privilege was set, single queue runner processes started
2565 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2566 same command in the process it spins off) were not dropping privilege.
2568 6. When the daemon running as "exim" started a queue runner, it always
2569 re-executed Exim in the spun-off process. This is a waste of effort when
2570 deliver_drop_privilege is set. The new process now just calls the
2571 queue-runner function directly.
2577 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2580 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2581 ignored. Though the use of -oP was forcing the writing of a pid file, it
2582 was always written to the default place.
2584 3. If the message "no IP address found for host xxxx" is generated during
2585 incoming verification, it is now followed by identification of the incoming
2586 connection (so you can more easily find what provoked it).
2588 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2590 5. Added some features to "harden" Exim a bit more against certain attacks:
2592 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2593 be put in Local/Makefile. This is like the never_users runtime option,
2594 but it cannot be overridden. The default setting is "root".
2596 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2597 prefix string with which any file named in a -C command line option
2600 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2601 is retained for -C and -D only if the caller of Exim is root. Without
2602 it, the exim user may also use -C and -D and retain privilege.
2604 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2605 command line option is disabled.
2607 6. Macro names set by the -D option must start with an upper case letter, just
2608 like macro names defined in the configuration file.
2610 7. Added "dereference=" facility to LDAP.
2612 8. Two instances of the typo "uknown" in the source files are fixed.
2614 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2615 the Configure-Makefile script screwed up while processing it.
2617 10. Incorporated PCRE 4.4.
2619 11. The SMTP synchronization check was not operating right at the start of an
2620 SMTP session. For example, it could not catch a HELO sent before the client
2621 waited for the greeting. There is now a check for outstanding input at the
2622 point when the greeting is written. Because of the duplex, asynchronous
2623 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2624 way, but not yet received, when the check is performed.
2626 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2627 on TCP/IP sockets, because this apparently causes some broken clients to
2630 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2631 unchanged) from the Cygwin maintainer.
2633 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2634 was supported instead of testing for mbx. Effectively a typo.
2636 15. The spa authenticator server code was not checking that the input it
2637 received was valid base64.
2639 16. The debug output line for the "set" modifier in ACLs was not showing the
2640 name of the variable that was being set.
2642 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2644 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2645 lookup on its own. Something else was needed to trigger the lookup. For
2646 example, a match in host_lookup or the need for a host name in a host list.
2647 Now, if $sender_host_name is referenced and the host name has not yet been
2648 looked up, a lookup is performed. If the lookup fails, the variable remains
2649 empty, and $host_lookup_failed is set to "1".
2651 19. Added "eqi" as a case-independent comparison operator.
2653 20. The saslauthd authentication condition could segfault if neither service
2654 nor realm was specified.
2656 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2657 error message that was logged was misleading, and incoming SMTP
2658 connections were dropped. The message is now more accurate, and temporary
2659 errors are given to SMTP connections.
2661 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2662 (except RSET) with a 421 error, until QUIT is received. However, it was
2663 failing to send a response to QUIT.
2665 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2666 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2667 ACL is now run after the helo_try_verify_hosts code.
2669 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2670 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2671 case-independent, so other case variants are also recognized). Apparently
2672 some systems use these upper case variants.
2674 25. If more than two messages were waiting for the same host, and a transport
2675 filter was specified for the transport, Exim sent two messages over the
2676 same TCP/IP connection, and then failed with "socket operation on non-
2677 socket" when it tried to send the third.
2679 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2681 27. The extern definition of crypt16() in expand.c was not being excluded when
2682 the OS had its own crypt16() function.
2684 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2685 as a preferred synonym for return_size_limit, both as an option and as an
2688 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2690 30. Changed the default configuration ACL to relax the local part checking rule
2691 for addresses that are not in any local domains. For these addresses,
2692 slashes and pipe symbols are allowed within local parts, but the sequence
2693 /../ is explicitly forbidden.
2695 31. SPA server authentication was not clearing the challenge buffer before
2698 32. log_message in a "warn" ACL statement was writing to the reject log as
2699 well as to the main log, which contradicts the documentation and doesn't
2700 seem right (because no rejection is happening). So I have stopped it.
2702 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2703 However, I am unable to do any testing of this.
2705 34. Fixed an infelicity in the appendfile transport. When checking directories
2706 for a mailbox, to see if any needed to be created, it was accidentally
2707 using path names with one or more superfluous leading slashes; tracing
2708 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2710 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2711 contents are added to the log line that is written for every discarded
2712 recipient. (Previously a log_message setting was ignored.)
2714 36. The ${quote: operator now quotes the string if it is empty.
2716 37. The install script runs exim in order to find its version number. If for
2717 some reason other than non-existence or emptiness, which it checks, it
2718 could not run './exim', it was installing it with an empty version number,
2719 i.e. as "exim-". This error state is now caught, and the installation is
2722 38. An argument was missing from the function that creates an error message
2723 when Exim fails to connect to the socket for saslauthd authentication.
2724 This could cause Exim to crash, or give a corrupted message.
2726 39. Added isip, isip4, and isip6 to ${if conditions.
2728 40. The ACL variables $acl_xx are now saved with the message, and can be
2729 accessed later in routers, transports, and filters.
2731 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2732 strings in the file are not string-expanded.
2734 42. If a MAIL command specified a SIZE value that was too large to fit into an
2735 int variable, the check against message_size_limit failed. Such values are
2736 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2737 day this will have to be increased, but I don't think I want to be around
2738 when emails are that large.
2745 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2746 iconv() is not standard in FreeBSD.
2748 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2749 IPv6 enabled. The observed symptom was a segmentation fault on return from
2750 the function os_common_find_running_interfaces() in src/os.c.
2752 3. In the check_special_case() function in daemon.c I had used "errno" as an
2753 argument name, which causes warnings on some systems. This was basically a
2754 typo, since it was named "eno" in the comments!
2756 4. The code that waits for the clock to tick (at a resolution of some fraction
2757 of a second) so as to ensure message-id uniqueness was always waiting for
2758 at least one whole tick, when it could have waited for less. [This is
2759 almost certainly not relevant at current processor speeds, where it is
2760 unlikely to ever wait at all. But we try to future-proof.]
2762 5. The function that sleeps for a time interval that includes fractions of a
2763 second contained a race. It did not block SIGALRM between setting the
2764 timer, and suspending (a couple of lines later). If the interval was short
2765 and the sigsuspend() was delayed until after it had expired, the suspension
2766 never ended. On busy systems this could lead to processes getting stuck for
2769 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2770 process, before it forks any delivery processes. The open lookup caching
2771 mechanism meant that the open file or database connection was passed into
2772 the delivery process. The problem was that delivery processes always tidy
2773 up cached lookup data. This could cause a problem for the next delivery
2774 process started by the queue runner, because the external queue runner
2775 process does not know about the closure. So the next delivery process
2776 still has data in the lookup cache. In the case of a file lookup, there was
2777 no problem because closing a file descriptor in a subprocess doesn't affect
2778 the parent. However, if the lookup was caching a connection to a database,
2779 the connection was closed, and the second delivery process was likely to
2780 see errors such as "PGSQL: query failed: server closed the connection
2781 unexpectedly". The problem has been fixed by closing all cached lookups
2782 in a queue runner before running a delivery process.
2784 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2785 seem to have the "const" qualifier which it has on other OS. I've
2788 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2789 *of the same type* (client or server) with the same public name that an
2790 error should be diagnosed.
2792 9. When Exim looked up a host name for an IP address, but failed to find the
2793 original IP address when looking up the host name (a safety check), it
2794 output the message "<ip address> does not match any IP for NULL", which was
2795 confusing, to say the least. The bug was that the host name should have
2796 appeared instead of "NULL".
2798 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2799 user that is built into the binary, and the -C or -D options is used, root
2800 privilege is dropped before the configuration file is read. In addition,
2801 logging is switched to stderr instead of the normal log files. If the
2802 configuration then re-defines the Exim user, the unprivileged environment
2803 is probably not what is expected, so Exim logs a panic warning message (but
2806 However, if deliver_drop_privilege is set, the unprivileged state may well
2807 be exactly what is intended, so the warning has been cut out in that case,
2808 and Exim is allowed to try to write to its normal log files.
2814 1. smtp_return_error_details was not giving details for temporary sender
2815 or receiver verification errors.
2817 2. Diagnose a configuration error if two authenticators have the same public
2820 3. Exim used not to create the message log file for a message until the first
2821 delivery attempt. This could be confusing when incoming messages were held
2822 for policy or load reasons. The message log file is now created at the time
2823 the message is received, and an initial "Received" line is written to it.
2825 4. The automatically generated man page for command line options had a minor
2826 bug that caused no ill effects; however, a more serious problem was that
2827 the procedure for building the man page automatically didn't always
2828 operate. Consequently, release 4.20 contains an out-of-date version. This
2829 shouldn't happen again.
2831 5. When building Exim with embedded Perl support, the script that builds the
2832 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2833 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2835 6. The freeze_tell option was not being used for messages that were frozen on
2836 arrival, either by an ACL or by local_scan().
2838 7. Added the smtp_incomplete_transaction log selector.
2840 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2841 was accepting AUTH without a new EHLO.
2843 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2844 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2847 10. Logging of TCP/IP connections (when configured) now happens in the main
2848 daemon process instead of the child process, so that the TCP/IP connection
2849 count is more accurate (but it can never be perfect).
2851 11. The use of "drop" in a nested ACL was not being handled correctly in the
2852 outer ACL. Now, if condition failure induced by the nested "drop" causes
2853 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2854 or "require"), the connection is dropped.
2856 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2857 that yield "discard" can now be used with an "accept" or a "discard" verb,
2858 but an error is generated for any others (because I can't see a useful way
2859 to define what should happen).
2861 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2862 are now processed in the same way as lines in the Exim configuration file.
2863 In particular, continuation lines are supported.
2865 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2867 15. Added -ti meaning -t -i.
2869 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2870 domains, and warn by logging if others are found.
2872 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2873 returned from GIFCONF when Exim is trying to find the list of interfaces on
2874 a host. The code in os.c has been modified to copy each ifreq to an aligned
2875 structure in all cases.
2877 Also, in some cases, the returned ifreq's were being copied to a 'struct
2878 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2879 means the last couple of bytes of an IPv6 address could be chopped if the
2880 ifreq contained only a normal sockaddr (14 bytes storage).
2882 18. Named domain lists were not supported in the hosts_treat_as_local option.
2883 An entry such as +xxxx was not recognized, and was treated as a literal
2886 19. Ensure that header lines added by a DATA ACL are included in the reject log
2887 if the ACL subsequently rejects the message.
2889 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2890 MD5 (which is deprecated).
2892 21. When testing a filter file using -bf, Exim was writing a message when it
2893 took the sender from a "From " line in the message, but it was not doing so
2894 when it took $return_path from a Return-Path: header line. It now does.
2896 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2897 with a valid header line field name (a series of printing characters
2898 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2900 23. Changed "disc" in the source to "disk" to conform to the documentation and
2901 the book and for uniformity.
2903 24. Ignore Sendmail's -Ooption=value command line item.
2905 25. When execve() failed while trying to run a command in a pipe transport,
2906 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2907 could be confused with a return value of 69 from the command itself. This
2908 has been changed to 127, the value the shell returns if it is asked to run
2909 a non-existent command. The wording for the related log line suggests a
2910 non-existent command as the problem.
2912 26. If received_header_text expands to an empty string, do not add a Received:
2913 header line to the message. (Well, it adds a token one on the spool, but
2914 marks it "old" so that it doesn't get used or transmitted.)
2916 27. Installed eximstats 1.28 (addition of -nt option).
2918 28. There was no check for failure on the call to getsockname() in the daemon
2919 code. This can fail if there is a shortage of resources on the system, with
2920 ENOMEM, for example. A temporary error is now given on failure.
2922 29. Contrary to the C standard, it seems that in some environments, the
2923 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2924 program. Exim now does this explicitly; it affects the formatting of
2925 timestamps using strftime().
2927 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2928 complaints. I've now initialized all the variables, to avoid this.
2930 32. Header lines added by a system filter were not being "seen" during
2931 transport-time rewrites.
2933 33. The info_callback() function passed to OpenSSL is set up with type void
2934 (*)(SSL *, int, int), as described somewhere. However, when calling the
2935 function (actually a macro) that sets it up, the type void(*)() is
2936 expected. I've put in a cast to prevent warnings from picky compilers.
2938 34. If a DNS black list lookup found a CNAME record, but there were no A
2939 records associated with the domain it pointed at, Exim crashed.
2941 35. If a DNS black list lookup returned more than one A record, Exim ignored
2942 all but the first. It now scans all returned addresses if a particular IP
2943 value is being sought. In this situation, the contents of the
2944 $dnslist_value variable are a list of all the addresses, separated by a
2947 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2948 to accept a host name and all its aliases if the forward lookup for any of
2949 them yielded the IP address of the incoming connection. Now it accepts only
2950 those names whose forward lookup yields the correct IP address. Any other
2951 names are discarded. This closes a loophole whereby a rogue DNS
2952 administrator could create reverse DNS records to break through a
2953 wildcarded host restriction in an ACL.
2955 37. If a user filter or a system filter that ran in a subprocess used any of
2956 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2957 the wrong values were passed to the pipe command ($thisaddress had the
2958 value of $0, $0 had the value of $1, etc). This bug was introduced by
2959 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2961 38. Improved the line breaking for long SMTP error messages from ACLs.
2962 Previously, if there was no break point between 40 and 75 characters, Exim
2963 left the rest of the message alone. Two changes have been made: (a) I've
2964 reduced the minimum length to 35 characters; (b) if it can't find a break
2965 point between 35 and 75 characters, it looks ahead and uses the first one
2966 that it finds. This may give the occasional overlong line, but at least the
2967 remaining text gets split now.
2969 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2970 file descriptors might be low, and that setting 1000 would always raise it.
2971 It turns out that in some environments, the limit is already over 1000 and
2972 that lowering it causes trouble. So now Exim takes care not to decrease it.
2974 40. When delivering a message, the value of $return_path is set to $sender_
2975 address at the start of routing (routers may change the value). By an
2976 oversight, this default was not being set up when an address was tested by
2977 -bt or -bv, which affected the outcome if any router or filter referred to
2980 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2981 log only when "message" or "log_message" are set. However, if one of the
2982 conditions was an address verification, or a call to a nested ACL, the
2983 messages generated by the underlying test were being passed through. This
2984 no longer happens. The underlying message is available in $acl_verify_
2985 message for both "message" and "log_message" expansions, so it can be
2986 passed through if needed.
2988 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2989 new expansion $bh_ to give the encoded byte string without charset
2990 translation. Translation happens only if iconv() is available; HAVE_ICONV
2991 indicates this at build time. HEADERS_CHARSET gives the charset to
2992 translate to; headers_charset can change it in the configuration, and
2993 "headers charset" can change it in an individual filter file.
2995 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
2996 that creates RFC 2047 encoded "words" labels them as that charset instead
2997 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
2998 expansion operator; (ii) when Exim creates a From: line for a local
2999 message; (iii) when a header line is rewritten to include a "phrase" part.
3001 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3002 buggy, causing it to skip the first lines of messages whose message ID
3003 ended in 'D'. This would not have bitten before Exim release 4.14, because
3004 message IDs were unlikely to end in 'D' before then. The effect was to have
3005 incorrect size information for certain domains.
3007 45. #include "config.h" was missing at the start of the crypt16.c module. This
3008 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3011 46. If there was a timeout during a "random" callout check, Exim treated it as
3012 a failure of the random address, and carried on sending RSET and the real
3013 address. If the delay was just some slowness somewhere, the response to the
3014 original RCPT would be taken as a response to RSET and so on, causing
3015 mayhem of various kinds.
3017 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3018 when I implemented it. It didn't allow for the fact that some option values
3019 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3020 the right test for positive values.
3022 48. Domain names in DNS records are case-independent. Exim always looks them up
3023 in lower case. Some resolvers return domain names in exactly the case they
3024 appear in the zone file, that is, they may contain uppercase letters. Not
3025 all resolvers do this - some return always lower case. Exim was treating a
3026 change of case by a resolver as a change of domain, similar to a widening
3027 of a domain abbreviation. This triggered its re-routing code and so it was
3028 trying to route what was effectively the same domain again. This normally
3029 caused routing to fail (because the router wouldn't handle the domain
3030 twice). Now Exim checks for this case specially, and just changes the
3031 casing of the domain that it ultimately uses when it transmits the message
3034 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3037 50. If a filter generated a file delivery with a non-absolute name (possible if
3038 no home directory exists for the router), the forbid_file option was not
3041 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3042 the existing equality matching.
3044 52. Exim was using ints instead of ino_t variables in some places where it was
3045 dealing with inode numbers.
3047 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3048 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3049 called TMPDIR, and if it finds it is different, it changes its value.
3051 54. The smtp_printf() function is now made available to local_scan() so
3052 additional output lines can be written before returning. There is also an
3053 smtp_fflush() function to enable the detection of a dropped connection.
3054 The variables smtp_input and smtp_batched_input are exported to
3057 55. Changed the default runtime configuration: the message "Unknown user"
3058 has been removed from the ACL, and instead placed on the localuser router,
3059 using the cannot_route_message feature. This means that any verification
3060 failures that generate their own messages won't get overridden. Similarly,
3061 the "Unrouteable address" message that was in the ACL for unverifiable
3062 relay addresses has also been removed.
3064 56. Added hosts_avoid_esmtp to the smtp transport.
3066 57. The exicyclog script was not checking for the esoteric option
3067 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3068 will work only if exicyclog is run under the appropriate euid.
3070 58. Following a discussion on the list, the rules by which Exim recognises line
3071 endings on incoming messages have been changed. The -dropcr and drop_cr
3072 options are now no-ops, retained only for backwards compatibility. The
3073 following line terminators are recognized: LF CRLF CR. However, special
3074 processing applies to CR:
3076 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3077 nor a local message in the state where . is a terminator.
3079 (ii) If a bare CR is encountered in a header line, an extra space is added
3080 after the line terminator so as not to end the header. The reasoning
3081 behind this is that bare CRs in header lines are most likely either
3082 to be mistakes, or people trying to play silly games.
3084 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3085 was being incorrectly given as 18 bytes larger than it should have been.
3086 This is a VOB (very old bug).
3088 60. This may never have affected anything current, but just in case it has:
3089 When the local host is found other than at the start of a list of hosts,
3090 the local host, those with the same MX, and any that follow, are discarded.
3091 When the list in question was part of a longer list of hosts, the following
3092 hosts (not currently being processed) were also being discarded. This no
3093 longer happens. I'm not sure if this situation could ever has previously
3096 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3099 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3100 header. This is something that is recommended in a new Internet Draft, and
3101 is something that is documented as being done by Sendmail. There are two
3102 possible values. For messages generated by the autoreply transport, Exim
3105 Auto-Submitted: auto-replied
3107 whereas for all other generated messages (e.g. bounces) it adds
3109 Auto-Submitted: auto-generated
3111 63. The "personal" condition in filters now includes a test for the
3112 Auto-Submitted: header. If it contains the string "auto-" the message it
3113 not considered personal.
3115 64. Added rcpt_include_affixes as a generic transport option.
3117 65. Added queue_only_override (default true).
3119 66. Added the syslog_duplication option.
3121 67. If what should have been the first header line of a message consisted of
3122 a space followed by a colon, Exim was mis-interpreting it as a header line.
3123 It isn't of course - it is syntactically invalid and should therefore be
3124 treated as the start of the message body. The misbehaviour could have
3125 caused a number of strange effects, including loss of data in subsequent
3126 header lines, and spool format errors.
3128 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3129 client host had authenticated. This control can now be exercised by an ACL
3130 for more flexibility.
3132 69. By default, callouts do not happen when testing with -bh. There is now a
3133 variant, -bhc, which does actually run the callout code, including
3134 consulting and updating the callout cache.
3136 70. Added support for saslauthd authentication, courtesy of Alexander
3139 71. If statvfs() failed on the spool or log directories while checking their
3140 size for availability, Exim confusingly gave the error "space shortage".
3141 Furthermore, in debugging mode it crashed with a floating point exception.
3142 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3143 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3144 serious problem, Exim now writes to the main and panic logs when this
3145 happens, with details of the failure. It then refuses to accept the
3146 incoming message, giving the message "spool directory problem" or "log
3147 directory problem" with a 421 code for SMTP messages.
3149 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3150 0, 1, and 2 exist, because some OS complain for execs without them (see
3151 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3152 descriptors. However, the code omitted to check that the open succeeded,
3153 causing mysterious errors if for some reason the permissions on /dev/null
3154 got screwed. Now Exim writes a message to the main and panic logs, and
3155 bombs out if it can't open /dev/null.
3157 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3158 interact so that it is all more flexible. It is supposed to remain
3159 backwards compatible. Also added extra_local_interfaces.
3161 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3162 to bomb out with an assertion failure - to the client this appears as a
3163 connection drop. This problem occurs in the part of the code that was taken
3164 from the Samba project. Fortunately, the assertion is in a very simple
3165 function, so I have fixed this by reproducing the function inline in the
3166 one place where it is called, and arranging for authentication to fail
3167 instead of killing the process with assert().
3169 75. The SPA client code was not working when the server requested OEM rather
3170 than Unicode encoding.
3172 76. Added code to make require_files with a specific uid setting more usable in
3173 the case where statting the file as root fails - usually a non-root-mounted
3174 NFS file system. When this happens and the failure is EACCES, Exim now
3175 forks a subprocess and does the per-uid checking as the relevant uid.
3177 77. Added process_log_path.
3179 78. If log_file_path was not explicitly set, a setting of check_log_space or
3180 check_log_inodes was ignored.
3182 79. If a space check for the spool or log partitions fails, the incident is now
3183 logged. Of course, in the latter case the data may get lost...
3185 80. Added the %p formatting code to string_format() so that it can be used to
3186 print addresses in debug_print(). Adjusted all the address printing in the
3187 debugging in store.c to use %p rather than %d.
3189 81. There was a concern that a line of code in smtp_in.c could overflow a
3190 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3191 initially expressed, the concern was not well-founded, because trailing
3192 spaces are removed early. However, if the trailing spaces were followed by
3193 a NULL, they did not get removed, so the overflow was possible. Two fixes
3196 (a) I re-wrote the offending code in a cleaner fashion.
3197 (b) If an incoming SMTP command contains a NULL character, it is rejected
3200 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3201 runs initgroups(), so that if the Exim user is in any additional groups,
3202 they will be used during message reception.
3208 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob