1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.209 2005/08/23 08:46:33 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Added the "success_on_redirect" address verification option. See
10 NewStuff for rationale and an example.
12 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
14 PH/02 Patch to exigrep to allow it to work on syslog lines.
16 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
17 fread() to skip over the body file's header line, because in Cygwin the
18 header line is locked and is inaccessible.
20 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
21 co-exist for some time) to make it clear that it is the Exim ID that is
22 referenced, not the Message-ID: header line.
24 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
25 string_format() function, because snprintf() does not exist on all
28 PH/06 The use of forbid_filter_existstest now also locks out the use of the
29 ${stat: expansion item.
31 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
32 protocol synchronization error", to keep the pedants happy.
34 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
35 well as for IRIX systems, when gcc is being used. See the host.c source
38 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
40 PH/10 Named domain lists were not working if used in a queue_smtp_domains
43 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
44 transport and to the smtp transport in LMTP mode.
46 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
48 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
49 run a filter in a subprocess. This could lead to confusion in subsequent
50 lookups in the parent process. There should also be a search_tidyup() at
51 the end of the subprocess.
53 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
54 only if the host matched helo_try_verify_hosts, which caused the
55 verification to occur when the EHLO/HELO command was issued. The ACL just
56 tested the remembered result. Now, if a previous verification attempt has
57 not happened, "verify = helo" does it there and then.
59 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
61 TK/03 Fix log output including CR from clamd.
63 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
64 exist provoked a memory error which could cause a segfault.
66 PH/15 Installed PCRE 6.2
68 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
70 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
71 of the problem. Specifically, suggested +O2 rather than +O1 for the
74 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
76 PH/20 If a delivery was routed to a non-standard port by means of an SRV
77 record, the port was not correctly logged when the outgoing_port log
78 selector was set (it logged the transort's default port).
80 PH/21 Added support for host-specific ports to manualroute, queryprogram,
81 fallback_hosts, and "hosts" in the smtp transport.
83 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
84 host errors such as "Connection refused".
86 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
87 authentication with radiusclient 0.4.9:
89 - Error returned from rc_read_config was caught wrongly
90 - Username/password not passed on to radius server due to wrong length.
92 The presumption is that some radiusclient API changes for 4.51/PH/17
93 were not taken care of correctly. The code is still untested by me (my
94 Linux distribution still has 0.3.2 of radiusclient), but it was
95 contributed by a Radius user.
97 PH/24 When doing a callout, the value of $domain wasn't set correctly when
98 expanding the "port" option of the smtp transport.
100 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
101 while reading a MIME header. Thanks to Tom Hughes for a patch.
103 PH/24 Include config.h inside local_scan.h so that configuration settings are
106 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
107 that in an ACL for RCPT (for example), you can examine exactly what was
110 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
111 commands, but it was not correctly comparing the address with the actual
112 client host address. Thus, it would show the EHLO address in Received:
113 header lines when this was not necessary.
115 PH/27 Added the % operator to ${eval:}.
121 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
123 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
124 command itself ended in error, the underlying message about the transport
125 filter timeout was being overwritten with the pipe command error. Now the
126 underlying error message should be appended to the second error message.
128 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
130 PH/02 Reduce the amount of output that "make" produces by default. Full output
131 can still be requested.
133 PH/03 The warning log line about a condition test deferring for a "warn" verb
134 was being output only once per connection, rather than after each
135 occurrence (because it was using the same function as for successful
136 "warn" verbs). This seems wrong, so I have changed it.
138 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
139 it should not have, which might have caused a crash in the right
140 circumstances, but probably never did.
142 PH/04 Installed a modified version of Tony Finch's patch to make submission
143 mode fix the return path as well as the Sender: header line, and to
144 add a /name= option so that you can make the user's friendly name appear
147 TF/03 Added the control = fakedefer ACL modifier.
149 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
150 Mark Lowes for thorough testing.
152 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
154 TK/03 Merged latest SRS patch from Miles Wilton.
156 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
157 with the definition in sysexits.h (which is #included earlier).
158 Fortunately, Exim does not actually use EX_OK. The code used to try to
159 preserve the sysexits.h value, by assumimg that macro definitions were
160 scanned for macro replacements. I have been disabused of this notion,
161 so now the code just undefines EX_OK before #including unistd.h.
163 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
164 in the smtp transport. When a block could not be written in a single
165 write() function, the timeout was being re-applied to each part-write.
166 This seems wrong - if the receiver was accepting one byte at a time it
167 would take for ever. The timeout is now adjusted when this happens. It
168 doesn't have to be particularly precise.
170 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
171 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
173 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
174 without a domain if the check to <postmaster@domain> fails.
176 SC/01 Eximstats: added -xls and the ability to specify output files
177 (patch written by Frank Heydlauf).
179 SC/02 Eximstats: use FileHandles for outputing results.
181 SC/03 Eximstats: allow any combination of xls, txt, and html output.
183 SC/04 Eximstats: fixed display of large numbers with -nvr option
185 SC/05 Eximstats: fixed merging of reports with empty tables.
187 SC/06 Eximstats: added the -include_original_destination flag
189 SC/07 Eximstats: removed tabs and trailing whitespace.
191 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
193 TK/06 MBOX spool code: Add real "From " MBOX separator line
194 so the .eml file is really in mbox format (even though
195 most programs do not really care). Patch from Alex Miller.
197 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
198 The latter is generated from $received_to and is only set if the
199 message has one envelope recipient. SA can use these headers,
200 obviously out-of-the-box. Patch from Alex Miller.
202 PH/08 The ${def test on a variable was returning false if the variable's
203 value was "0", contrary to what the specification has always said!
204 The result should be true unless the variable is empty.
206 PH/09 The syntax error of a character other than { following "${if
207 def:variable_name" (after optional whitespace) was not being diagnosed.
208 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
209 accidental colon was present, for example, could give incorrect results.
211 PH/10 Tidied the code in a number of places where the st_size field of a stat()
212 result is used (not including appendfile, where other changes are about
215 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
216 This involved changing a lot of size variables from int to off_t. It
217 should work with maildirs and everything.
219 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
220 spamd dying while we are connected to it.
222 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
225 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
226 function for BDB 4.3.
228 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
230 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
231 the sources for PCRE have been re-arranged and more files are now
234 PH/15 The code I had for printing potentially long long variables in PH/11
235 above was not the best (it lost precision). The length of off_t variables
236 is now inspected at build time, and an appropriate printing format (%ld
237 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
238 to be "long long int" or "long int". This is needed for the internal
239 formatting function string_vformat().
241 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
242 the configuration file to be ":syslog", then the script "guesses" where
243 the logs files are, rather than using the compiled in default. In our
244 case the guess is not the same as the compiled default, so the script
245 suddenly stopped working when I started to use syslog. The patch checks
246 to see if log_file_path is "". If so, it attempts to read it from exim
247 with no configuration file to get the compiled in version, before it
248 falls back to the previous guessing code."
250 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
251 implementing BATV in an Exim configuration. See NewStuff for the gory
254 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
255 Makefile that are specific to HP-UX.
257 PH/18 If the "use_postmaster" option was set for a recipient callout together
258 with the "random" option, the postmaster address was used as the MAIL
259 FROM address for the random test, but not for the subsequent recipient
260 test. It is now used for both.
262 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
263 patch removes a few documentation additions to RFC 3028, because the
264 latest draft now contains them. It adds the new en;ascii-case comparator
265 and a new error check for 8bit text in MIME parts. Comparator and
266 require names are now matched exactly. I enabled the subaddress
267 extension, but it is not well tested yet (read: it works for me)."
269 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
270 rework some of the code of TK/09 above to avoid the hardwired use of
271 "%lld" and "long long". Replaced the call to snprintf() with a call to
274 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
275 records point to non-existent hosts", "retry timeout exceeded", and
276 "retry time not reached for any host after a long failure period".
278 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
279 experimental DomainKeys support:
281 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
282 (2) On an error such as an illegally used "control", the wrong name for
283 the control was given.
285 These problems did NOT occur unless DomainKeys support was compiled.
287 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
289 PH/24 Added ${if match_ip condition.
291 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
292 databases so that it will be absolutely obvious if a crash occurs in the
293 DB library. This is a regular occurrence (often caused by mis-matched
296 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
297 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
298 picked up on a user's system that detects such things. There doesn't seem
299 to be a gcc warning option for this - only an attribute that has to be
300 put on the function's prototype. It seems that in Fedora Core 4 they have
301 set this on a number of new functions. No doubt there will be more in due
304 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
305 specify a transport. However, if an address that was verified by such a
306 router was the subject of a callout, Exim crashed because it tried to
307 read the rcpt_include_affixes from the non-existent transport. Now it
308 just assumes that the setting of that option is false. This bug was
309 introduced by 4.51/PH/31.
311 PH/28 Changed -d+all to exclude +memory, because that information is very
312 rarely of interest, but it makes the output a lot bigger. People tend to
313 do -d+all out of habit.
315 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
316 code in os-type was giving problems when libc.so lives in lib64, like on
319 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
320 aren't the modern standard, and it seems that some systems' include files
321 don't always have them. Exim was already checking for some of the newer
322 ones like T_AAAA, and defining it itself. I've added checks for all the
323 record types that Exim uses.
325 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
326 not automatically generating a new one, as it is supposed to. This
327 prevented TLS from working. If the file did exist, but contained invalid
328 data, a new version was generated, as expected. It was only the case of a
329 non-existent file that was broken.
331 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
332 with a change in libdomainkeys > 0.64.
334 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
335 from DNS. If the selector record carries the flag, it now has
336 precedence over the domain-wide flag.
338 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
340 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
341 the use of a port name that isn't defined in /etc/services) occurred, the
342 message was deferred as in a normal delivery, and thus remained on the
343 spool, instead of being failed because of the mua_wrapper setting. This
344 is now fixed, and I tidied up some of the mua_wrapper messages at the
347 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
348 the messages in a hash of arrays rather than using individual hashes.
349 This is a bit cleaner and results in dramatic memory savings, albeit
350 at a slight CPU cost.
352 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
353 as requested by Marc Sherman.
355 SC/10 Eximstats: added histograms for user specified patterns as requested
358 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
360 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
361 fopen() in the content-scanning modules that did not already have it.
367 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
368 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
370 TK/02 Fix ACL "control" statement not being available in MIME ACL.
372 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
374 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
375 to test Sieve filters that use "vacation".
377 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
378 that changes the way the GnuTLS parameters are stored in the cache file.
379 The new format can be generated externally. For backward compatibility,
380 if the data in the cache doesn't make sense, Exim assumes it has read an
381 old-format file, and it generates new data and writes a new file. This
382 means that you can't go back to an older release without removing the
385 PH/03 A redirect router that has both "unseen" and "one_time" set does not
386 work if there are any delivery delays because "one_time" forces the
387 parent to be marked "delivered", so its unseen clone is never tried
388 again. For this reason, Exim now forbids the simultaneous setting of
391 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
392 redirected to themselves ("homonym" addresses). Read the long ChangeLog
393 entry if you want to know the details. The fix, however, neglected to
394 consider the case when local delivery batching is involved. The test for
395 "previously delivered" was not happening when checking to see if an
396 address could be batched with a previous (undelivered) one; under
397 certain circumstances this could lead to multiple deliveries to the same
400 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
401 in its include files, and this causes problems building Exim.
403 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
404 header_syntax) but Exim was just ignoring anything given after a slash.
405 In particular, this caused confusion with an attempt to use "verify =
406 reverse_host_lookup/defer_ok". An error is now given when options are
407 supplied for verify items that do not have them. (Maybe reverse_host_
408 lookup should have a defer_ok option, but that's a different point.)
410 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
411 defined by RFC 821) to 2048, because there were problems with some AUTH
412 commands, and RFC 1869 says the size should be increased for extended
413 SMTP commands that take arguments.
415 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
418 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
419 "unknown" error; now it says that the functionality isn't in the binary.
421 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
422 an address' error message when a string expansion fails (syntax or
423 whatever). Otherwise the password may appear in the log. Following change
424 PH/42 below, there is no longer a chance of it appearing in a bounce
427 PH/11 Installed exipick version 20050225.0 from John Jetmore.
429 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
430 of its addresses was ever tried. (Bugzilla bug #2.)
432 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
433 the result incorrectly in the debug output. (It correctly added a newline
434 to what was transported.)
436 TF/01 Added $received_time.
438 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
439 commented out examples of how to interface to a virus scanner and to
440 SpamAssassin. Also added commented examples of av_scanner and
441 spamd_address settings.
443 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
444 and controls are allowed in which ACLs. There were a couple of minor
445 errors. Some of the entries in the conditions table (which is a table of
446 where they are NOT allowed) were getting very unwieldy; rewrote them as a
447 negation of where the condition IS allowed.
449 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
451 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
452 header file does not have a version number, so I've had to invent a new
453 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
454 API. The code is untested by me (my Linux distribution still has 0.3.2 of
455 radiusclient), but it was contributed by a Radius user.
457 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
458 files or directories, for OpenSSL.
460 PH/19 When an Exim process that is running as root has to create an Exim log
461 file, it does so in a subprocess that runs as exim:exim so as to get the
462 ownership right at creation (otherwise, other Exim processes might see
463 the file with the wrong ownership). There was no test for failure of this
464 fork() call, which would lead to the process getting stuck as it waited
465 for a non-existent subprocess. Forks do occasionally fail when resources
466 run out. I reviewed all the other calls to fork(); they all seem to check
469 PH/20 When checking for unexpected SMTP input at connect time (before writing
470 the banner), Exim was not dealing correctly with a non-positive return
471 from the read() function. If the client had disconnected by this time,
472 the result was a log entry for a synchronization error with an empty
473 string after "input=" when read() returned zero. If read() returned -1
474 (an event I could not check), uninitialized data bytes were printed.
475 There were reports of junk text (parts of files, etc) appearing after
478 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
480 PH/22 Added support for macro redefinition, and (re)definition in between
481 driver and ACL definitions.
483 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
484 forgetting to use the resulting value; it was using the unexpanded value.
486 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
487 hadn't been configured. The fix is from Juergen Kreileder, who
488 understands it better than I do:
490 "Here's what I see happening with three configured cyrus_sasl
491 authenticators configured (plain, login, cram-md5):
493 On startup auth_cyrus_sasl_init() gets called for each of these.
494 This means three calls to sasl_listmech() without a specified mech_list.
495 => SASL tests which mechs of all available mechs actually work
496 => three warnings about OTP not working
497 => the returned list contains: plain, login, cram-md5, digest-md5, ...
499 With the patch, sasl_listmech() also gets called three times. But now
500 SASL's mech_list option is set to the server_mech specified in the the
501 authenticator. Or in other words, the answer from sasl_listmech()
502 gets limited to just the mech you're testing for (which is different
504 => the return list contains just 'plain' or 'login', 'cram-md5' or
505 nothing depending on the value of ob->server_mech.
507 I've just tested the patch: Authentication still works fine,
508 unavailable mechs specified in the exim configuration are still
509 caught, and the auth.log warnings about OTP are gone."
511 PH/25 When debugging is enabled, the contents of the command line are added
512 to the debugging output, even when log_selector=+arguments is not
515 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
516 answer is "GNU", and only if the return is "GNU/something" is the answer
519 PH/27 $acl_verify_message is now set immediately after the failure of a
520 verification in an ACL, and so is available in subsequent modifiers. In
521 particular, the message can be preserved by coding like this:
523 warn !verify = sender
524 set acl_m0 = $acl_verify_message
526 Previously, $acl_verify_message was set only while expanding "message"
527 and "log_message" when a very denied access.
529 PH/28 Modified OS/os.c-Linux with
531 -#ifndef OS_LOAD_AVERAGE
532 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
534 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
535 nomenclature these days.)
537 PH/29 Installed patch from the Sieve maintainer that adds the options
538 sieve_useraddress and sieve_subaddress to the redirect router.
540 PH/30 In these circumstances:
541 . Two addresses routed to the same list of hosts;
542 . First host does not offer TLS;
543 . First host accepts first address;
544 . First host gives temporary error to second address;
545 . Second host offers TLS and a TLS session is established;
546 . Second host accepts second address.
547 Exim incorrectly logged both deliveries with the TLS parameters (cipher
548 and peerdn, if requested) that were in fact used only for the second
551 PH/31 When doing a callout as part of verifying an address, Exim was not paying
552 attention to any local part prefix or suffix that was matched by the
553 router that accepted the address. It now behaves in the same way as it
554 does for delivery: the affixes are removed from the local part unless
555 rcpt_include_affixes is set on the transport.
557 PH/32 Add the sender address, as F=<...>, to the log line when logging a
558 timeout during the DATA phase of an incoming message.
560 PH/33 Sieve envelope tests were broken for match types other than :is. I have
561 applied a patch sanctioned by the Sieve maintainer.
563 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
564 the uid or gid is negative. A case of a negative gid caused this to be
565 noticed. The fix allows for either to be negative.
567 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
568 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
571 PH/36 The change PH/12 above was broken. Fixed it.
573 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
574 the grounds that routing the same address twice would always produce the
575 same answer. This might have been true once, but it is certainly no
576 longer true now. Routing a child address may depend on the previous
577 routing that produced that child. Some complicated redirection strategies
578 went wrong when messages had multiple recipients, and made Exim's
579 behaviour dependent on the order in which the addresses were given.
581 I have moved the duplicate checking until after the routing is complete.
582 Exim scans the addresses that are assigned to local and remote
583 transports, and removes any duplicates. This means that more work will be
584 done, as duplicates will always all be routed, but duplicates are
585 presumably rare, so I don't expect this is of any significance.
587 For deliveries to pipes, files, and autoreplies, the duplicate checking
588 still happens during the routing process, since they are not going to be
591 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
592 It corrects a timeout issue with spamd. This is Ian's comment: "The
593 background is that sometimes spamd either never reads data from a
594 connection it has accepted, or it never writes response data. The exiscan
595 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
596 blindly assumes that writes won't block so it may never time out."
598 PH/39 Allow G after quota size as well as K and M.
600 PH/40 The value set for $authenticated_id in an authenticator may not contain
601 binary zeroes or newlines because the value is written to log lines and
602 to spool files. There was no check on this. Now the value is run through
603 the string_printing() function so that such characters are converted to
604 printable escape sequences.
606 PH/41 $message_linecount is a new variable that contains the total number of
607 lines in the message. Compare $body_linecount, which is the count for the
610 PH/42 Exim no longer gives details of delivery errors for specific addresses in
611 bounce and delay warning messages, except in certain special cases, which
614 (a) An SMTP error message from a remote host;
615 (b) A message specified in a :fail: redirection;
616 (c) A message specified in a "fail" command in a system filter;
617 (d) A message specified in a FAIL return from the queryprogram router;
618 (e) A message specified by the cannot_route_message router option.
620 In these cases only, Exim does include the error details in bounce and
621 warning messages. There are also a few cases where bland messages such
622 as "unrouteable address" or "local delivery error" are given.
624 PH/43 $value is now also set for the "else" part of a ${run expansion.
626 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
627 being worked on, but at least Exim now implements the latest version to
630 PH/45 In a pipe transport, although a timeout while waiting for the pipe
631 process to complete was treated as a delivery failure, a timeout while
632 writing the message to the pipe was logged, but erroneously treated as a
633 successful delivery. Such timeouts include transport filter timeouts. For
634 consistency with the overall process timeout, these timeouts are now
635 treated as errors, giving rise to delivery failures by default. However,
636 there is now a new Boolean option for the pipe transport called
637 timeout_defer, which, if set TRUE, converts the failures into defers for
638 both kinds of timeout. A transport filter timeout is now identified in
641 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
642 systems where "make" and "gmake" are different, calling "gmake" at top
643 level broke things. I've arranged for the value of $(MAKE) to be passed
644 from the Makefile to this script so that it can call the same version of
648 A note about Exim versions 4.44 and 4.50
649 ----------------------------------------
651 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
652 changes of various kinds. As a consequence, a big documentation update was
653 needed. This delayed the release for rather longer than seemed good, especially
654 in the light of a couple of (minor) security issues. Therefore, the changes
655 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
656 release. So 4.44 and 4.50 are in effect two different branches that both start
659 I have left the 4.50 change log unchanged; it contains all the changes since
660 4.43. The change log for 4.44 is below; many of its items are identical to
661 those for 4.50. This seems to be the most sensible way to preserve the
662 historical information.
668 1. Minor wording change to the doc/README.SIEVE file.
670 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
671 computation of the current number of files was incorrect.
673 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
674 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
677 4. Give more explanation in the error message when the command for a transport
678 filter fails to execute.
680 5. There are several places where Exim runs a non-Exim command in a
681 subprocess. The SIGUSR1 signal should be disabled for these processes. This
682 was being done only for the command run by the queryprogram router. It is
683 now done for all such subprocesses. The other cases are: ${run, transport
684 filters, and the commands run by the lmtp and pipe transports.
686 6. Added CONFIGURE_GROUP build-time option.
688 7. Some older OS have a limit of 256 on the maximum number of file
689 descriptors. Exim was using setrlimit() to set 1000 as a large value
690 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
691 systems. I've change it so that if it can't get 1000, it tries for 256.
693 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
694 was an oversight, and furthermore, ever since the addition of extra
695 controls (e.g. 4.43/32), the checks on when to allow different forms of
696 "control" were broken. There should now be diagnostics for all cases when a
697 control that does not make sense is encountered.
699 9. Added the /retain_sender option to "control=submission".
701 10. $recipients is now available in the predata ACL (oversight).
703 11. Tidy the search cache before the fork to do a delivery from a message
704 received from the command line. Otherwise the child will trigger a lookup
705 failure and thereby defer the delivery if it tries to use (for example) a
706 cached ldap connection that the parent has called unbind on.
708 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
709 of $address_data from the recipient verification was clobbered by the
712 13. The value of address_data from a sender verification is now available in
713 $sender_address_data in subsequent conditions in the ACL statement.
715 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
717 15. Added a new option "connect=<time>" to callout options, to set a different
720 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
721 was its contents. (It was OK if the option was not defined at all.)
723 17. A "Completed" log line is now written for messages that are removed from
724 the spool by the -Mrm option.
726 18. New variables $sender_verify_failure and $recipient_verify_failure contain
727 information about exactly what failed.
729 19. Added -dd to debug only the daemon process.
731 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
732 handles timeouts, both on the server side and network timeouts. Renamed the
733 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
735 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
736 It is now set to "smtps".
738 22. $host_address is now set to the target address during the checking of
741 23. When checking ignore_target_hosts for an ipliteral router, no host name was
742 being passed; this would have caused $sender_host_name to have been used if
743 matching the list had actually called for a host name (not very likely,
744 since this list is usually IP addresses). A host name is now passed as
747 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
748 code that specifies a non-restarting handler (typically sigaction() in
749 modern systems) in an attempt to fix a rare and obscure crash bug.
751 25. Narrowed the window for a race in the daemon that could cause it to ignore
752 SIGCHLD signals. This is not a major problem, because they are used only to
753 wake it up if nothing else does.
755 26. A malformed maildirsize file could cause Exim to calculate negative values
756 for the mailbox size or file count. Odd effects could occur as a result.
757 The maildirsize information is now recalculated if the size or filecount
760 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
761 support for a long time. Removed HAVE_SYS_VFS_H.
763 28. Installed the latest version of exipick from John Jetmore.
765 29. In an address list, if the pattern was not a regular expression, an empty
766 subject address (from a bounce message) matched only if the pattern was an
767 empty string. Non-empty patterns were not even tested. This was the wrong
768 because it is perfectly reasonable to use an empty address as part of a
769 database query. An empty address is now tested by patterns that are
770 lookups. However, all the other forms of pattern expect the subject to
771 contain a local part and a domain, and therefore, for them, an empty
772 address still always fails if the pattern is not itself empty.
774 30. Exim went into a mad DNS loop when attempting to do a callout where the
775 host was specified on an smtp transport, and looking it up yielded more
778 31. Re-factored the code for checking spool and log partition space into a
779 function that finds that data and another that does the check. The former
780 is then used to implement four new variables: $spool_space, $log_space,
781 $spool_inodes, and $log_inodes.
783 32. The RFC2047 encoding function was originally intended for short strings
784 such as real names; it was not keeping to the 75-character limit for
785 encoded words that the RFC imposes. It now respects the limit, and
786 generates multiple encoded words if necessary. To be on the safe side, I
787 have increased the buffer size for the ${rfc2047: expansion operator from
790 33. It is now permitted to omit both strings after an "if" condition; if the
791 condition is true, the result is "true". As before, when the second string
792 is omitted, a false condition yields an empty string. This makes it less
793 cumbersome to write custom ACL and router conditions.
795 34. Failure to deliver a bounce message always caused it to be frozen, even if
796 there was an errors_to setting on the router. The errors_to setting is now
799 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
800 canonical form (fully expanded) before being placed in
801 $sender_host_address.
803 36. The table in the code that translates DNS record types into text (T_A to
804 "A" for instance) was missing entries for NS and CNAME. It is just possible
805 that this could have caused confusion if both these types were looked up
806 for the same domain, because the text type is used as part of Exim's
807 per-process caching. But the chance of anyone hitting this buglet seems
810 37. The dnsdb lookup has been extended in a number of ways.
812 (1) There is a new type, "zns", which walks up the domain tree until it
813 finds some nameserver records. It should be used with care.
815 (2) There is a new type, "mxh", which is like "mx" except that it returns
816 just the host names, not the priorities.
818 (3) It is now possible to give a list of domains (or IP addresses) to be
819 looked up. The behaviour when one of the lookups defers can be
820 controlled by a keyword.
822 (4) It is now possible to specify the separator character for use when
823 multiple records are returned.
825 38. The dnslists ACL condition has been extended: it is now possible to supply
826 a list of IP addresses and/or domains to be looked up in a particular DNS
829 39. Added log_selector=+queue_time_overall.
831 40. When running the queue in the test harness, wait just a tad after forking a
832 delivery process, to get repeatability of debugging output.
834 41. Include certificate and key file names in error message when GnuTLS fails
835 to set them up, because the GnuTLS error message doesn't include the name
836 of the failing file when there is a problem reading it.
838 42. Allow both -bf and -bF in the same test run.
840 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
842 44. The "Exiscan patch" is now merged into the mainline Exim source.
844 45. Sometimes the final signoff response after QUIT could fail to get
845 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
846 < 0 before doing a fflush(). This bug looks as though it goes back to the
847 introduction of TLS in release 3.20, but "sometimes" must have been rare
848 because the tests only now provoked it.
850 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
851 (this can affect the format of dates).
853 47. exim_tidydb, when checking for the continued existence of a message for
854 which it has found a message-specific retry record, was not finding
855 messages that were in split spool directories. Consequently, it was
856 deleting retry records that should have stayed in existence.
858 48. Steve fixed some bugs in eximstats.
860 49. The SPA authentication driver was not abandoning authentication and moving
861 on to the next authenticator when an expansion was forced to fail,
862 contradicting the general specification for all authenticators. Instead it
863 was generating a temporary error. It now behaves as specified.
865 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
866 (the order specifies the preference for clients). The order is now AES256,
867 AES128, 3DES, ARCFOUR128.
869 51. Small patch to Sieve code - explicitly set From: when generating an
872 52. Exim crashed if a remote delivery caused a very long error message to be
873 recorded - for instance if somebody sent an entire SpamAssassin report back
874 as a large number of 550 error lines. This bug was coincidentally fixed by
875 increasing the size of one of Exim's internal buffers (big_buffer) that
876 happened as part of the Exiscan merge. However, to be on the safe side, I
877 have made the code more robust (and fixed the comments that describe what
880 53. Now that there can be additional text after "Completed" in log lines (if
881 the queue_time_overall log selector is set), a one-byte patch to exigrep
882 was needed to allow it to recognize "Completed" as not the last thing in
885 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
886 patch that reportedly fixes this has been added. I am not expert enough to
887 create a test for it. This is what the patch creator wrote:
889 "I found a little strange behaviour of ldap code when working with
890 Windows 2003 AD Domain, where users was placed in more than one
891 Organization Units. When I tried to give exim partial DN, the exit code
892 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
893 But simultaneously result of request was absolutely normal ldap result,
894 so I produce this patch..."
896 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
897 REFERENCE, so I have modified the code to exclude the patch when that macro
900 55. Some experimental protocols are using DNS PTR records for new purposes. The
901 keys for these records are domain names, not reversed IP addresses. The
902 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
903 leaves it alone. Component reversal etc. now happens only for IP addresses.
906 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
908 57. Double the size of the debug message buffer (to 2048) so that more of very
909 long debug lines gets shown.
911 58. The exicyclog utility now does better if the number of log files to keep
912 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
914 59. Two changes related to the smtp_active_hostname option:
916 (1) $smtp_active_hostname is now available as a variable.
917 (2) The default for smtp_banner uses $smtp_active_hostname instead
918 of $primary_hostname.
920 60. The host_aton() function is supposed to be passed a string that is known
921 to be a valid IP address. However, in the case of IPv6 addresses, it was
922 not checking this. This is a hostage to fortune. Exim now panics and dies
923 if the condition is not met. A case was found where this could be provoked
924 from a dnsdb PTR lookup with an IPv6 address that had more than 8
925 components; fortuitously, this particular loophole had already been fixed
926 by change 4.50/55 above.
928 If there are any other similar loopholes, the new check in host_aton()
929 itself should stop them being exploited. The report I received stated that
930 data on the command line could provoke the exploit when Exim was running as
931 exim, but did not say which command line option was involved. All I could
932 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
936 61. There was a buffer overflow vulnerability in the SPA authentication code
937 (which came originally from the Samba project). I have added a test to the
938 spa_base64_to_bits() function which I hope fixes it.
941 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
942 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
944 63. The daemon start-up calls getloadavg() while still root for those OS that
945 need the first call to be done as root, but it missed one case: when
946 deliver_queue_load_max is set with deliver_drop_privilege. This is
947 necessary for the benefit of the queue runner, because there is no re-exec
948 when deliver_drop_privilege is set.
950 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
953 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
954 was in use, was not putting the data itself into the right store pool;
955 consequently, it could be overwritten for a subsequent message in the same
956 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
959 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
961 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
962 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
963 the function were treating the return as a boolean value, which happened to
964 work because 0=false and not-0=true, but is not correct code.
966 68. The host_aton() function was not handling scoped IPv6 addresses (those
967 with, for example, "%eth0" on the end) correctly.
969 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
970 negated items (that is, ~something) in unsigned ints. Some compilers
971 apparently mutter when there is no cast.
973 70. If an address verification called from an ACL failed, and did not produce a
974 user-specific message (i.e. there was only a "system" message), nothing was
975 put in $acl_verify_message. In this situation, it now puts the system
978 71. Change 4.23/11 added synchronization checking at the start of an SMTP
979 session; change 4.31/43 added the unwanted input to the log line - except
980 that it did not do this in the start of session case. It now does.
982 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
983 This is wrong and can cause the other end to generate a synchronization
984 error if it is another Exim or anything else that does the synchronization
985 check. A QUIT command is no longer sent after a timeout.
987 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
990 74. The defer_ok option of callout verification was not working if it was used
991 when verifying addresses in header lines, that is, for this case:
993 verify = header_sender/callout=defer_ok
995 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
996 those file descriptors could be used for SMTP connections. If anything
997 wrote to stderr (the example that came up was "warn" in embedded Perl), it
998 could be sent to the SMTP client, causing chaos. The daemon now opens
999 stdin, stdout, and stderr to /dev/null when it puts itself into the
1002 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1003 log by default. The user can override this with suitable Perl magic.
1005 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1006 the log message when discard triggers, was not working for the DATA ACL or
1007 for the non-SMTP ACL.
1009 78. Error message wording change in sieve.c.
1011 79. If smtp_accept_max_per_host was set, the number of connections could be
1012 restricted to fewer than expected, because the daemon was trying to set up
1013 a new connection before checking whether the processes handling previous
1014 connections had finished. The check for completed processes is now done
1015 earlier. On busy systems, this bug wouldn't be noticed because something
1016 else would have woken the daemon, and it would have reaped the completed
1019 80. If a message was submitted locally by a user whose login name contained one
1020 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1021 It caused a spool format error. I have fixed the spool reading code. A
1022 related problem was that the "from" clause in the Received: line became
1023 illegal because of the space(s). It is now covered by ${quote_local_part.
1025 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1028 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1030 83. Patch from Sieve maintainer to fix unterminated string problem in
1031 "vacation" handling.
1033 84. Some minor changes to the Linux configuration files to help with other
1034 OS variants using glibc.
1036 85. One more patch for Sieve to update vacation handling to latest spec.
1039 ----------------------------------------------------
1040 See the note above about the 4.44 and 4.50 releases.
1041 ----------------------------------------------------
1047 1. Change 4.43/35 introduced a bug that caused file counts to be
1048 incorrectly computed when quota_filecount was set in an appendfile
1051 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1052 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1055 3. Give more explanation in the error message when the command for a transport
1056 filter fails to execute.
1058 4. There are several places where Exim runs a non-Exim command in a
1059 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1060 was being done only for the command run by the queryprogram router. It is
1061 now done for all such subprocesses. The other cases are: ${run, transport
1062 filters, and the commands run by the lmtp and pipe transports.
1064 5. Some older OS have a limit of 256 on the maximum number of file
1065 descriptors. Exim was using setrlimit() to set 1000 as a large value
1066 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1067 systems. I've change it so that if it can't get 1000, it tries for 256.
1069 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1070 was an oversight, and furthermore, ever since the addition of extra
1071 controls (e.g. 4.43/32), the checks on when to allow different forms of
1072 "control" were broken. There should now be diagnostics for all cases when a
1073 control that does not make sense is encountered.
1075 7. $recipients is now available in the predata ACL (oversight).
1077 8. Tidy the search cache before the fork to do a delivery from a message
1078 received from the command line. Otherwise the child will trigger a lookup
1079 failure and thereby defer the delivery if it tries to use (for example) a
1080 cached ldap connection that the parent has called unbind on.
1082 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1083 of $address_data from the recipient verification was clobbered by the
1084 sender verification.
1086 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1087 was its contents. (It was OK if the option was not defined at all.)
1089 11. A "Completed" log line is now written for messages that are removed from
1090 the spool by the -Mrm option.
1092 12. $host_address is now set to the target address during the checking of
1093 ignore_target_hosts.
1095 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1096 being passed; this would have caused $sender_host_name to have been used if
1097 matching the list had actually called for a host name (not very likely,
1098 since this list is usually IP addresses). A host name is now passed as
1101 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1102 code that specifies a non-restarting handler (typically sigaction() in
1103 modern systems) in an attempt to fix a rare and obscure crash bug.
1105 15. Narrowed the window for a race in the daemon that could cause it to ignore
1106 SIGCHLD signals. This is not a major problem, because they are used only to
1107 wake it up if nothing else does.
1109 16. A malformed maildirsize file could cause Exim to calculate negative values
1110 for the mailbox size or file count. Odd effects could occur as a result.
1111 The maildirsize information is now recalculated if the size or filecount
1114 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1115 support for a long time. Removed HAVE_SYS_VFS_H.
1117 18. Updated exipick to current release from John Jetmore.
1119 19. Allow an empty sender to be matched against a lookup in an address list.
1120 Previously the only cases considered were a regular expression, or an
1123 20. Exim went into a mad DNS lookup loop when doing a callout where the
1124 host was specified on the transport, if the DNS lookup yielded more than
1127 21. The RFC2047 encoding function was originally intended for short strings
1128 such as real names; it was not keeping to the 75-character limit for
1129 encoded words that the RFC imposes. It now respects the limit, and
1130 generates multiple encoded words if necessary. To be on the safe side, I
1131 have increased the buffer size for the ${rfc2047: expansion operator from
1134 22. Failure to deliver a bounce message always caused it to be frozen, even if
1135 there was an errors_to setting on the router. The errors_to setting is now
1138 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1139 canonical form (fully expanded) before being placed in
1140 $sender_host_address.
1142 24. Updated eximstats to version 1.33
1144 25. Include certificate and key file names in error message when GnuTLS fails
1145 to set them up, because the GnuTLS error message doesn't include the name
1146 of the failing file when there is a problem reading it.
1148 26. Expand error message when OpenSSL has problems setting up cert/key files.
1151 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1152 (this can affect the format of dates).
1154 28. exim_tidydb, when checking for the continued existence of a message for
1155 which it has found a message-specific retry record, was not finding
1156 messages that were in split spool directories. Consequently, it was
1157 deleting retry records that should have stayed in existence.
1159 29. eximstats updated to version 1.35
1160 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1161 1.35 - bugfix such that pie charts by volume are generated correctly
1163 30. The SPA authentication driver was not abandoning authentication and moving
1164 on to the next authenticator when an expansion was forced to fail,
1165 contradicting the general specification for all authenticators. Instead it
1166 was generating a temporary error. It now behaves as specified.
1168 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1169 (the order specifies the preference for clients). The order is now AES256,
1170 AES128, 3DES, ARCFOUR128.
1172 31. Small patch to Sieve code - explicitly set From: when generating an
1175 32. Exim crashed if a remote delivery caused a very long error message to be
1176 recorded - for instance if somebody sent an entire SpamAssassin report back
1177 as a large number of 550 error lines. This bug was coincidentally fixed by
1178 increasing the size of one of Exim's internal buffers (big_buffer) that
1179 happened as part of the Exiscan merge. However, to be on the safe side, I
1180 have made the code more robust (and fixed the comments that describe what
1183 33. Some experimental protocols are using DNS PTR records for new purposes. The
1184 keys for these records are domain names, not reversed IP addresses. The
1185 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1186 leaves it alone. Component reversal etc. now happens only for IP addresses.
1189 34. The host_aton() function is supposed to be passed a string that is known
1190 to be a valid IP address. However, in the case of IPv6 addresses, it was
1191 not checking this. This is a hostage to fortune. Exim now panics and dies
1192 if the condition is not met. A case was found where this could be provoked
1193 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1194 components; fortuitously, this particular loophole had already been fixed
1195 by change 4.50/55 or 4.44/33 above.
1197 If there are any other similar loopholes, the new check in host_aton()
1198 itself should stop them being exploited. The report I received stated that
1199 data on the command line could provoke the exploit when Exim was running as
1200 exim, but did not say which command line option was involved. All I could
1201 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1202 running as the user.
1205 35. There was a buffer overflow vulnerability in the SPA authentication code
1206 (which came originally from the Samba project). I have added a test to the
1207 spa_base64_to_bits() function which I hope fixes it.
1210 36. The daemon start-up calls getloadavg() while still root for those OS that
1211 need the first call to be done as root, but it missed one case: when
1212 deliver_queue_load_max is set with deliver_drop_privilege. This is
1213 necessary for the benefit of the queue runner, because there is no re-exec
1214 when deliver_drop_privilege is set.
1216 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1217 was in use, was not putting the data itself into the right store pool;
1218 consequently, it could be overwritten for a subsequent message in the same
1219 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1222 38. Sometimes the final signoff response after QUIT could fail to get
1223 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1224 < 0 before doing a fflush(). This bug looks as though it goes back to the
1225 introduction of TLS in release 3.20, but "sometimes" must have been rare
1226 because the tests only now provoked it.
1232 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1233 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1234 Now it returns an integer. A number of calls were still expecting a T/F
1235 return. Fortuitously, in all cases, the tests worked in OK situations,
1236 which is the norm. However, things would have gone wrong on any write
1237 failures on the smtp file descriptor. This function is used when sending
1238 messages over SMTP and also when doing verify callouts.
1240 2. When Exim is called to do synchronous delivery of a locally submitted
1241 message (the -odf or -odi options), it no longer closes stderr before doing
1244 3. Implemented the mua_wrapper option.
1246 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1248 5. Implemented the functions header_remove(), header_testname(),
1249 header_add_at_position(), and receive_remove_recipient(), and exported them
1252 6. If an ACL "warn" statement specified the addition of headers, Exim already
1253 inserted X-ACL-Warn: at the start if there was no header name. However, it
1254 was not making this test for the second and subsequent header lines if
1255 there were newlines in the string. This meant that an invalid header could
1256 be inserted if Exim was badly configured.
1258 7. Allow an ACL "warn" statement to add header lines at the start or after all
1259 the Received: headers, as well as at the end.
1261 8. Added the rcpt_4xx retry error code.
1263 9. Added postmaster_mailfrom=xxx to callout verification option.
1265 10. Added mailfrom=xxxx to the callout verification option, for verify=
1268 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1269 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1271 12. Inserted some casts to stop certain compilers warning when using pointer
1272 differences as field lengths or precisions in printf-type calls (mostly
1273 affecting debugging statements).
1275 13. Added optional readline() support for -be (dynamically loaded).
1277 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1278 same clock tick as a message's arrival, so that its received time was the
1279 same as the "first fail" time on the retry record, and that message
1280 remained on the queue past the ultimate address timeout, every queue runner
1281 would try a delivery (because it was past the ultimate address timeout) but
1282 after another failure, the ultimate address timeout, which should have then
1283 bounced the address, did not kick in. This was a "< instead of <=" error;
1284 in most cases the first failure would have been in the next clock tick
1285 after the received time, and all would be well.
1287 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1288 being recognized when the domain list was tested by the match_domain
1289 condition in an expansion string.
1291 16. Added the ${str2b64: operator.
1293 17. Exim was always calling setrlimit() to set a large limit for the number of
1294 processes, without checking whether the existing limit was already
1295 adequate. (It did check for the limit on file descriptors.) Furthermore,
1296 errors from getrlimit() and setrlimit() were being ignored. Now they are
1297 logged to the main and panic logs, but Exim does carry on, to try to do its
1298 job under whatever limits there are.
1300 18. Imported PCRE 5.0.
1302 19. Trivial typo in log message " temporarily refused connection" (the leading
1305 20. If the log selector return_path_on_delivery was set and an address was
1306 redirected to /dev/null, the delivery process crashed because it assumed
1307 that a return path would always be set for a "successful" delivery. In this
1308 case, the whole delivery is bypassed as an optimization, and therefore no
1311 21. Internal re-arrangement: the function for sending a challenge and reading
1312 a response while authentication was assuming a zero-terminated challenge
1313 string. It's now changed to take a pointer and a length, to allow for
1314 binary data in such strings.
1316 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1318 23. Exim was not respecting finduser_retries when seeking the login of the
1319 uid under which it was called; it was always trying 10 times. (The default
1320 setting of finduser_retries is zero.) Also, it was sleeping after the final
1321 failure, which is pointless.
1323 24. Implemented tls_on_connect_ports.
1325 25. Implemented acl_smtp_predata.
1327 26. If the domain in control=submission is set empty, Exim assumes that the
1328 authenticated id is a complete email address when it generates From: or
1329 Sender: header lines.
1331 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1332 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1333 chown and chgrp in /bin and hostname in /usr/bin.
1335 28. Exim was keeping the "process log" file open after each use, just as it
1336 does for the main log. This opens the possibility of it remaining open for
1337 long periods when the USR1 signal hits a daemon. Occasional processlog
1338 errors were reported, that could have been caused by this. Anyway, it seems
1339 much more sensible not to leave this file open at all, so that is what now
1342 29. The long-running daemon process does not normally write to the log once it
1343 has entered its main loop, and it closes the log before doing so. This is
1344 so that log files can straightforwardly be renamed and moved. However,
1345 there are a couple of unusual error situations where the daemon does write
1346 log entries, and I had neglected to close the log afterwards.
1348 30. The text of an SMTP error response that was received during a remote
1349 delivery was being truncated at 512 bytes. This is too short for some of
1350 the long messages that one sometimes sees. I've increased the limit to
1353 31. It is now possible to make retry rules that apply only when a message has a
1354 specific sender, in particular, an empty sender.
1356 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1357 it possible to be selective about when SMTP synchronization is enforced.
1359 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1361 32. Implemented hosts_connection_nolog.
1363 33. Added an ACL for QUIT.
1365 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1368 35. Added mailbox_size and mailbox_filecount to appendfile.
1370 36. Added control = no_multiline_responses to ACLs.
1372 37. There was a bug in the logic of the code that waits for the clock to tick
1373 in the case where the clock went backwards by a substantial amount such
1374 that the microsecond fraction of "now" was more than the microsecond
1375 fraction of "then" (but the whole seconds number was less).
1377 38. Added support for the libradius Radius client library this is found on
1378 FreeBSD (previously only the radiusclient library was supported).
1384 1. When certain lookups returned multiple values in the form name=value, the
1385 quoting of the values was not always being done properly. Specifically:
1386 (a) If the value started with a double quote, but contained no whitespace,
1388 (b) If the value contained whitespace other than a space character (i.e.
1389 tabs or newlines or carriage returns) it was not quoted.
1390 This fix has been applied to the mysql and pgsql lookups by writing a
1391 separate quoting function and calling it from the lookup code. The fix
1392 should probably also be applied to nisplus, ibase and oracle lookups, but
1393 since I cannot test any of those, I have not disturbed their existing code.
1395 2. A hit in the callout cache for a specific address caused a log line with no
1396 reason for rejecting RCPT. Now it says "Previous (cached) callout
1397 verification failure".
1399 3. There was an off-by-one bug in the queryprogram router. An over-long
1400 return line was truncated at 256 instead of 255 characters, thereby
1401 overflowing its buffer with the terminating zero. As well as fixing this, I
1402 have increased the buffer size to 1024 (and made a note to document this).
1404 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1405 when Exim is waiting for an SMTP response from a remote server, Exim
1406 restarts its select() call on the socket, thereby resetting its timeout.
1407 This is not a problem when such interrupts are rare. Somebody set up a cron
1408 job to run exiwhat every 2 minutes, which is less than the normal select()
1409 timeout (5 or 10 minutes). This meant that the select() timeout never
1410 kicked in because it was always reset. I have fixed this by comparing the
1411 time when an interrupt arrives with the time at the start of the first call
1412 to select(). If more time than the timeout has elapsed, the interrupt is
1413 treated as a timeout.
1415 5. Some internal re-factoring in preparation for the addition of Sieve
1416 extensions (by MH). In particular, the "personal" test is moved to a
1417 separate function, and given an option for scanning Cc: and Bcc: (which is
1418 not set for Exim filters).
1420 6. When Exim created an email address using the login of the caller as the
1421 local part (e.g. when creating a From: or Sender: header line), it was not
1422 quoting the local part when it contained special characters such as @.
1424 7. Installed new OpenBSD configuration files.
1426 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1427 try to make them clearer.
1429 9. Callout options, other than the timeout value, were being ignored when
1430 verifying sender addresses in header lines. For example, when using
1432 verify = header_sender/callout=no_cache
1434 the cache was (incorrectly) being used.
1436 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1437 only the Cygwin environment.
1439 11. When return_path_on_delivery was set as a log selector, if different remote
1440 addresses in the same message used different return paths and parallel
1441 remote delivery occurred, the wrong values would sometimes be logged.
1442 (Whenever a remote delivery process finished, the return path value from
1443 the most recently started remote delivery process was logged.)
1445 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1446 lines when AUTH and/or TLS are in use. This is the "received protocol"
1447 field. Exim used to use "asmtp" for authenticated SMTP, without any
1448 indication (in the protocol name) for TLS use. Now it follows the RFC and
1449 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1450 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1451 names appear in log lines as well as in Received: header lines.
1453 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1454 extensions, and comparison tests, and to fix some bugs.
1456 14. Changes to the "personal" filter test:
1458 (1) The test was buggy in that it was just doing the equivalent of
1459 "contains" tests on header lines. For example, if a user's address was
1460 anne@some.where, the "personal" test would incorrectly be true for
1462 To: susanne@some.where
1464 This test is now done by extracting each address from the header in turn,
1465 and checking the entire address. Other tests that are part of "personal"
1466 are now done using regular expressions (for example, to check local parts
1467 of addresses in From: header lines).
1469 (2) The list of non-personal local parts in From: addresses has been
1470 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1471 taken from the Sieve specification recommendations.
1473 (3) If the message contains any header line starting with "List-" it is
1474 treated as non-personal.
1476 (4) The test for "circular" in the Subject: header line has been removed
1477 because it now seems ill-conceived.
1479 15. Minor typos in src/EDITME comments corrected.
1481 16. Installed latest exipick from John Jetmore.
1483 17. If headers_add on a router specified a text string that was too long for
1484 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1485 of string_sprintf() is now avoided.
1487 18. $message_body_size was not set (it was always zero) when running the DATA
1488 ACL and the local_scan() function.
1490 19. For the "mail" command in an Exim filter, no default was being set for
1491 the once_repeat time, causing a random time value to be used if "once" was
1492 specified. (If the value happened to be <= 0, no repeat happened.) The
1493 default is now 0s, meaning "never repeat". The "vacation" command was OK
1494 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1495 (I found it when inspecting the code).
1497 20. There is now an overall timeout for performing a callout verification. It
1498 defaults to 4 times the callout timeout, which applies to individual SMTP
1499 commands during the callout. The overall timeout applies when there is more
1500 than one host that can be tried. The timeout is checked before trying the
1501 next host. This prevents very long delays if there are a large number of
1502 hosts and all are timing out (e.g. when the network connections are timing
1503 out). The value of the overall timeout can be changed by specifying an
1504 additional sub-option for "callout", called "maxwait". For example:
1506 verify = sender/callout=5s,maxwait=20s
1508 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1509 to the end before writing, but this should make it even safer).
1511 22. Exim was forgetting that it had advertised PIPELINING for the second and
1512 subsequent messages on an SMTP connection. It was also not resetting its
1513 memory on STARTTLS and an internal HELO.
1515 23. When Exim logs an SMTP synchronization error within a session, it now
1516 records whether PIPELINING has been advertised or not.
1518 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1519 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1520 rather than long int.
1522 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1524 26. Added the never_mail option to autoreply.
1530 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1531 crash if the getsockname() call failed; this can happen if a connection is
1532 closed very soon after it is established. The problem was simply in the
1533 order in which certain operations were done, causing Exim to try to write
1534 to the SMTP stream before it had set up the file descriptor. The bug has
1535 been fixed by making things happen in the correct order.
1541 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1542 before the connection was closed, thus losing the rejection response.
1544 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1545 some early Solaris releases, but causes trouble in current releases where
1546 socklen_t is defined.
1548 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1551 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1554 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1555 that releasing the top of it at the end releases what was used for sub-
1556 expansions (unless the block got too big). However, discard this block if
1557 the first thing is a variable or header, so that we can use its block when
1558 it is dynamic (useful for very large $message_headers, for example).
1560 6. Lookups now cache *every* query, not just the most recent. A new, separate
1561 store pool is used for this. It can be recovered when all lookup caches are
1562 flushed. Lookups now release memory at the end of their result strings.
1563 This has involved some general refactoring of the lookup sources.
1565 7. Some code has been added to the store_xxx() functions to reduce the amount
1566 of flapping under certain conditions.
1568 8. log_incoming_interface used to affect only the <= reception log lines. Now
1569 it causes the local interface and port to be added to several more SMTP log
1570 lines, for example "SMTP connection from", and rejection lines.
1572 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1574 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1576 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1577 could be overwritten at the end of the current message (or the start of a
1578 new message if it was set in a HELO ACL). The value is now preserved for
1579 the duration of the SMTP connection.
1581 12. If a transport had a headers_rewrite setting, and a matching header line
1582 contained an unqualified address, that address was qualified, even if it
1583 did not match any rewriting rules. The underlying bug was that the values
1584 of the flags that permit the existence of unqualified sender and recipient
1585 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1586 non-local messages, and by -bnq for local messages) were not being
1587 preserved with the message after it was received.
1589 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1590 "next input=" as part of the text comprising the host identity instead of
1591 the correct text. The code was using the same buffer for two different
1592 strings. However, depending on which order the printing function evaluated
1593 its arguments, the bug did not always show up. Under Linux, for example, my
1594 test suite worked just fine.
1596 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1597 doesn't work with some older versions of Perl. It has been changed to "my",
1598 which in any case is probably the better facility to use.
1600 15. A really picky compiler found some instances of statements for creating
1601 error messages that either had too many or two few arguments for the format
1604 16. The size of the buffer for calls to the DNS resolver has been increased
1605 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1606 for addresses that have a lot of PTR records. This alleviates a problem; it
1607 does not fully solve it.
1609 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1610 buffer now truncates the list and logs the incident, which is the same
1611 action as happens when Exim is looking up a host name and its aliases.
1612 Previously in this situation something unpredictable would happen;
1613 sometimes it was "internal error: store_reset failed".
1615 18. If a server dropped the connection unexpectedly when an Exim client was
1616 using GnuTLS and trying to read a response, the client delivery process
1617 crashed while trying to generate an error log message.
1619 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1620 string, for example:
1622 warn message = H1: something\nH2: something
1624 the text was added as a single header line from Exim's point of view
1625 though it ended up OK in the delivered message. However, searching for the
1626 second and subsequent header lines using $h_h2: did not work. This has been
1627 fixed. Similarly, if a system filter added multiple headers in this way,
1628 the routers could not see them.
1630 20. Expanded the error message when iplsearch is called with an invalid key to
1631 suggest using net-iplsearch in a host list.
1633 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1634 ACLs are no longer actually imposed (and a message to that effect is
1637 22. If a "gecos" field in a passwd entry contained escaped characters, in
1638 particular, if it contained a \" sequence, Exim got it wrong when building
1639 a From: or a Sender: header from that name. A second bug also caused
1640 incorrect handling when an unquoted " was present following a character
1641 that needed quoting.
1643 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1644 was not being matched caselessly.
1646 24. Arranged for all hyphens in the exim.8 source to be escaped with
1649 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1650 properly. Recipient callout cache records were still being keyed to include
1651 the sender, even when use_sender was set false. This led to far more
1652 callouts that were necessary. The sender is no longer included in the key
1653 when use_sender is false.
1655 26. Added "control = submission" modifier to ACLs.
1657 27. Added the ${base62d: operator to decode base 62 numbers.
1659 28. dnsdb lookups can now access SRV records.
1661 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1662 the configuration file.
1664 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1665 (-v) mode. This makes the output for a verbose queue run more intelligible.
1667 31. Added a use_postmaster feature to recipient callouts.
1669 32. Added the $body_zerocount variable, containing the number of binary zero
1670 bytes in the message body.
1672 33. The time of last modification of the "new" subdirectory is now used as the
1673 "mailbox time last read" when there is a quota error for a maildir
1676 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1678 35. Added +ignore_unknown as a special item in host lists.
1680 36. Code for decoding IPv6 addresses in host lists is now included, even if
1681 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1682 address was recognized as an IP address, but was then not correctly decoded
1683 into binary, causing unexpected and incorrect effects when compared with
1690 1. Very minor rewording of debugging text in manualroute to say "list of
1691 hosts" instead of "hostlist".
1693 2. If verify=header_syntax was set, and a header line with an unqualified
1694 address (no domain) and a large number of spaces between the end of the
1695 name and the colon was received, the reception process suffered a buffer
1696 overflow, and (when I tested it) crashed. This was caused by some obsolete
1697 code that should have been removed. The fix is to remove it!
1699 3. When running in the test harness, delay a bit after writing a bounce
1700 message to get a bit more predictability in the log output.
1702 4. Added a call to search_tidyup() just before forking a reception process. In
1703 theory, someone could use a lookup in the expansion of smtp_accept_max_
1704 per_host which, without the tidyup, could leave open a database connection.
1706 5. Added the variables $recipient_data and $sender_data which get set from a
1707 lookup success in an ACL "recipients" or "senders" condition, or a router
1708 "senders" option, similar to $domain_data and $local_part_data.
1710 6. Moved the writing of debug_print from before to after the "senders" test
1713 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1714 problems for message scanning, either using a data ACL, or using
1715 local_scan() because the Received: header was not generated till after they
1716 were called (in order to set the time as the time of reception completion).
1717 I have revised the way this works. The header is now generated after the
1718 body is received, but before the ACL or local_scan() are called. After they
1719 are run, the timestamp in the header is updated.
1725 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1726 before starting a queue runner without re-exec. This happened only when
1727 deliver_drop_privilege was set or when the Exim user was set to root. The
1728 effect of the bug was that timeouts during subsequent deliveries caused
1729 crashes instead of being properly handled. The handler is now left at its
1730 default (and expected) setting.
1732 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1733 message, again when deliver_drop_privilege is set or Exim is run as root.
1734 The bug described in (1) was not present in this case, but the tidying up
1735 of the other signals was missing. I have made the two cases consistent.
1737 3. The ignore_target_hosts setting on a manualroute router was being ignored
1738 for hosts that were looked up using the /MX notation.
1740 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1743 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1744 operated on the sender address. After changing the $sender_address to <>
1745 for the sender address verify, Exim was re-instated it as the original
1746 (before rewriting) address, but remembering that it had rewritten it, so it
1747 wasn't rewriting it again. This bug also had the effect of breaking the
1748 sender address verification caching when the sender address was rewritten.
1750 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1751 This has been changed so that if the ip literal address matches
1752 ignore_target_hosts, the router declines.
1754 7. Added expansion conditions match_domain, match_address, and match_local_
1755 part (NOT match_host).
1757 8. The placeholder for the Received: header didn't have a length field set.
1759 9. Added code to Exim itself and to exim_lock to test for a specific race
1760 condition that could lead to file corruption when using MBX delivery. The
1761 issue is with the lockfile that is created in /tmp. If this file is removed
1762 after a process has opened it but before that process has acquired a lock,
1763 there is the potential for a second process to recreate the file and also
1764 acquire a lock. This could lead to two Exim processes writing to the file
1765 at the same time. The added code performs the same test as UW imapd; it
1766 checks after acquiring the lock that its file descriptor still refers to
1767 the same named file.
1769 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1770 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1773 11. Added the smtp_active_hostname option. If used, this will typically be made
1774 to depend on the incoming interface address. Because $interface_address is
1775 not set up until the daemon has forked a reception process, error responses
1776 that can happen earlier (such as "too many connections") no longer contain
1779 12. If an expansion in a condition on a "warn" statement fails because a lookup
1780 defers, the "warn" statement is abandoned, and the next ACL statement is
1781 processed. Previously this caused the whole ACL to be aborted.
1783 13. Added the iplsearch lookup type.
1785 14. Added ident_timeout as a log selector.
1787 15. Added tls_certificate_verified as a log selector.
1789 16. Added a global option tls_require_ciphers (compare the smtp transport
1790 option of the same name). This controls incoming TLS connections.
1792 17. I finally figured out how to make tls_require_ciphers do a similar thing
1793 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1794 before starting the TLS session.
1796 18. Tabs are now shown as \t in -bP output.
1798 19. If the log selector return_path_on_delivery was set, Exim crashed when
1799 bouncing a message because it had too many Received: header lines.
1801 20. If two routers both had headers_remove settings, and the first one included
1802 a superfluous trailing colon, the final name in the first list and the
1803 first name in the second list were incorrectly joined into one item (with a
1804 colon in the middle).
1810 1. Added -C and -D options to the exinext utility, mainly to make it easier
1811 to include in the automated testing, but these could be helpful when
1812 multiple configurations are in use.
1814 2. The exinext utility was not formatting the output nicely when there was
1815 an alternate port involved in the retry record key, nor when there was a
1816 message id as well (for retries that were specific to a specific message
1817 and a specific host). It was also confused by IPv6 addresses, because of
1818 the additional colons they contain. I have fixed the IPv4 problem, and
1819 patched it up to do a reasonable job for IPv6.
1821 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1822 delivery, the log line now contains "pipelined" if PIPELINING was used.
1824 4. An SMTP transport process used to panic and die if the bind() call to set
1825 an explicit outgoing interface failed. This has been changed; it is now
1826 treated in the same way as a connect() failure.
1828 5. A reference to $sender_host_name in the part of a conditional expansion
1829 that was being skipped was still causing a DNS lookup. This no longer
1832 6. The def: expansion condition was not recognizing references to header lines
1833 that used bh_ and bheader_.
1835 7. Added the _cache feature to named lists.
1837 8. The code for checking quota_filecount in the appendfile transport was
1838 allowing one more file than it should have been.
1840 9. For compatibility with Sendmail, the command line option
1848 and sets the incoming protocol and host name (for trusted callers). The
1849 host name and its colon can be omitted when only the protocol is to be set.
1850 Note the Exim already has two private options, -pd and -ps, that refer to
1851 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1852 "s", but I don't think that's a major issue.
1854 10. A number of refactoring changes to the code, none of which should affect
1857 (a) The number of logging options was getting close to filling up the
1858 32-bit word that was used as a bit map. I have split them into two classes:
1859 those that are passed in the argument to log_write(), and those that are
1860 only ever tested independently outside of that function. These are now in
1861 separate 32-bit words, so there is plenty of room for expansion again.
1862 There is no change in the user interface or the logging behaviour.
1864 (b) When building, for example, log lines, the code previously used a
1865 macro that called string_cat() twice, in order to add two strings. This is
1866 not really sufficiently general. Furthermore, there was one instance where
1867 it was actually wrong because one of the argument was used twice, and in
1868 one call a function was used. (As it happened, calling the function twice
1869 did not affect the overall behaviour.) The macro has been replaced by a
1870 function that can join an arbitrary number of extra strings onto a growing
1873 (c) The code for expansion conditions now uses a table and a binary chop
1874 instead of a serial search (which was left over from when there were very
1875 few conditions). Also, it now recognizes conditions like "pam" even when
1876 the relevant support is not compiled in: a suitably worded error message is
1877 given if an attempt is made to use such a condition.
1879 11. Added ${time_interval:xxxxx}.
1881 12. A bug was causing one of the ddress fields not to be passed back correctly
1882 from remote delivery subprocesses. The field in question was not being
1883 subsequently used, so this caused to problems in practice.
1885 13. Added new log selectors queue_time and deliver_time.
1887 14. Might have fixed a bug in maildirsizefile handling that threw up
1888 "unexpected character" debug warnings, and recalculated the data
1889 unnecessarily. In any case, I expanded the warning message to give more
1892 15. Added the message "Restricted characters in address" to the statements in
1893 the default ACL that block characters like @ and % in local parts.
1895 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1896 Three changes have been made:
1898 (a) There was a serious bug; a negative response to MAIL caused the whole
1899 recipient domain to be cached as invalid, thereby blocking all messages
1900 to all local parts at the same domain, from all senders. This bug has
1901 been fixed. The domain is no longer cached after a negative response to
1902 MAIL if the sender used is not empty.
1904 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1907 (c) A new callout option, "use_sender" has been added for people who want
1908 the modified behaviour.
1914 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1917 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1918 indeed breaks things for older releases.
1920 3. Added additional logging to the case where there is a problem reading data
1921 from a filter that is running in a subprocess using a pipe, in order to
1922 try to track down a specific problem.
1924 4. Testing facility fudge: when running in the test harness and attempting
1925 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1926 getting "No route to host". Convert this to a timeout.
1928 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1931 6. Some OS don't have socklen_t but use size_t instead. This affects the
1932 fifth argument of getsockopt() amongst other things. This is now
1933 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1934 can be set for individual OS. I have set it for SunOS5, OSF1, and
1935 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1936 some earlier ones do not.
1938 7. Change 4.30/15 was not doing the test caselessly.
1940 8. The standard form for an IPv6 address literal was being rejected by address
1941 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1942 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1943 this, as well as the form without the "IPv6" on the front (but only when
1944 address literals are enabled, of course).
1946 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1948 10. Exim crashed if a message with an empty sender address specified by -f
1949 encountered a router with an errors_to setting. This could be provoked only
1950 by a command such as
1954 where an empty string was supplied; "<>" did not hit this bug.
1956 11. Installed PCRE release 4.5.
1958 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1959 remained set. It is now erased.
1961 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1962 times from message ids (which are base 36 rather than the normal 62).
1964 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1965 were being counted as actual protocol errors, and logged if the log
1966 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1967 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1968 and DATA following a set of rejected RCPTs do not count as protocol errors.
1969 In other words, Exim assumes they were pipelined, though this may not
1970 actually be the case. Of course, in all cases the client gets an
1971 appropriate error code.
1973 15. If a lookup fails in an ACL condition, a message about the failure may
1974 be available; it is used if testing the ACL cannot continue, because most
1975 such messages specify what the cause of the deferral is. However, some
1976 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1977 that caused an old message to be retained and used if a later statement
1978 caused a defer, replacing the real cause of the deferral.
1980 16. If an IP address had so many PTR records that the DNS lookup buffer
1981 was not large enough to hold them, Exim could crash while trying to process
1982 the truncated data. It now detects and logs this case.
1984 17. Further to 4.21/58, another change has been made: if (and only if) the
1985 first line of a message (the first header line) ends with CRLF, a bare LF
1986 in a subsequent header line has a space inserted after it, so as not to
1987 terminate the header.
1989 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1990 unnecessarily, used atoi() instead of strtol(), and didn't check the
1991 termination when getting file sizes from file names by regex.
1993 19. Completely re-implemented the support for maildirsize files, in the light
1994 of a number of problems with the previous contributed implementation
1995 (4.30/29). In particular:
1997 . If the quota is zero, the maildirsize file is maintained, but no quota is
2000 . If the maildir directory does not exist, it is created before any attempt
2001 to write a maildirsize file.
2003 . The quota value in the file is just a cache; if the quota is changed in
2004 the transport, the new value overrides.
2006 . A regular expression is available for excluding directories from the
2009 20. The autoreply transport checks the characters in options that define the
2010 message's headers; it allows continued headers, but it was checking with
2011 isspace() after an embedded newline instead of explicitly looking for a
2014 21. If all the "regular" hosts to which an address was routed had passed their
2015 expiry times, and had not reached their retry times, the address was
2016 bounced, even if fallback hosts were defined. Now Exim should go on to try
2019 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2020 equivalent code in the SMTP transport. Some hosts send humungous responses
2021 to HELO/EHLO, more than 1024 it seems.
2023 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2024 gives compiler warnings in some environments. I've now done it "properly",
2027 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2028 (because of problems with the built-in one) was declared to return uschar *
2029 instead of char *, causing compiler failure.
2031 25. Fixed a file descriptor leak when processing alias/forward files.
2033 26. Fixed a minor format string issue in dbfn.c.
2035 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2037 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2038 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2041 29. When a sender address is verified, it is cached, to save repeating the test
2042 when there is more than one recipient in a message. However, when the
2043 verification involves a callout, it is possible for different callout
2044 options to be set for different recipients. It is too complicated to keep
2045 track of this in the cache, so now Exim always runs a verification when a
2046 callout is required, relying on the callout cache for the optimization.
2047 The overhead is duplication of the address routing, but this should not be
2050 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2051 to be verified with callout=postmaster, and the main callout worked but the
2052 postmaster check failed, the verification correctly failed. However, if a
2053 subsequent RCPT command asked for sender verification *without* the
2054 postmaster check, incorrect caching caused this verification also to fail,
2057 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2058 it was not caching the DNS options (qualify_single, search_parents) that
2059 were used when the lookup failed. A subsequent lookup with different
2060 options therefore always gave the same answer, though there were cases
2061 where it should not have. (Example: a "domains = !$mx_any" option on a
2062 dnslookup router: the "domains" option is always processed without any
2063 widening, but the router might have qualify_single set.) Now Exim uses the
2064 cached value only when the same options are set.
2066 32. Added John Jetmore's "exipick" utility to the distribution.
2068 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2069 than a timeout (e.g. a certificate is required, and is not provided), an
2070 Exim server now closes the connection immediately. Previously it waited for
2071 the client to close - but if the client is SSL, it seems that they each
2072 wait for each other, leading to a delay before one of them times out.
2074 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2075 maintained 0.8.x compatibility because I don't think many are using it, and
2076 it is clearly obsolete.
2078 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2081 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2082 client certificate was expired. A simple patch fixes this, though I don't
2083 understand the full logic of why the verify callback is called multiple
2086 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2087 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2088 which causes problems with some clients (such as the Certicom SSL Plus
2089 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2090 disables the coutermeasure allowing Eudora to connect."
2092 38. Exim was not checking that a write() to a log file succeeded. This could
2093 lead to Bad Things if a log got too big, in particular if it hit a file
2094 size limit. Exim now panics and dies if it cannot write to a log file, just
2095 as it does if it cannot open a log file.
2097 39. Modified OS/Makefile-Linux so that it now contains
2099 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2101 The two -D definitions ensure that Exim is compiled with large file
2102 support, which makes it possible to handle log files that are bigger than
2105 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2106 instance) a domain was checked against a named list that involved a lookup,
2107 causing $domain_data to be set, then another domain was checked against the
2108 same list, then the first domain was re-checked, the value of $domain_data
2109 after the final check could be wrong. In particular, if the second check
2110 failed, it could be set empty. This bug probably also applied to
2113 41. The strip_trailing_dot option was not being applied to the address given
2114 with the -f command-line option.
2116 42. The code for reading a message's header from the spool was incrementing
2117 $received_count, but never initializing it. This meant that the value was
2118 incorrect (doubled) while delivering a message in the same process in which
2119 it was received. In the most common configuration of Exim, this never
2120 happens - a fresh exec is done - but it can happen when
2121 deliver_drop_privilege is set.
2123 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2124 it now includes up to 150 characters of the unexpected data in the log
2127 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2128 and building data strings. The size of both of these buffers was 10 000
2129 bytes - far larger than anybody would *ever* want, thought I. Needless to
2130 say, somebody hit the limit. I have increased the maximum line length to
2131 20 000 and the maximum data length of concatenated lines to 100 000. I have
2132 also fixed two bugs, because there was no checking on these buffers. Tsk,
2133 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2134 buffer is too small.
2136 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2137 lsearch lookups. Now it does.
2139 46. When parsing a route_list item in a manualroute router, a fixed-length
2140 buffer was used for the list of hosts. I made this 1024 bytes long,
2141 thinking that nobody would ever have a list of hosts that long. Wrong.
2142 Somebody had a whole pile of complicated expansion conditions, and the
2143 string was silently truncated, leading to an expansion error. It turns out
2144 that it is easier to change to an unlimited length (owing to other changes
2145 that have happened since this code was originally written) than to build
2146 structure for giving a limitation error. The length of the item that
2147 expands into the list of hosts is now unlimited.
2149 47. The lsearch lookup could not handle data where the length of text line was
2150 more than 4095 characters. Such lines were truncated, leading to shortened
2151 data being returned. It should now handle lines of any length.
2153 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2154 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2157 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2158 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2159 no longer happens in comment lines. A list of replacements is now placed
2160 at the head of all of the source files, except those whose only change is
2161 to replace PERL_COMMAND in the very first #! line.
2163 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2164 messages on the queue, with a bottom-up merge sort, using code contributed
2165 by Michael Haardt. This should make operations like -bp somewhat faster on
2166 large queues. It won't affect queue runners, except when queue_run_in_order
2169 51. Installed eximstats 1.31 in the distribution.
2171 52. Added support for SRV lookups to the dnslookup router.
2173 53. If an ACL referred to $message_body or $message_body_end, the value was not
2174 reset for any messages that followed in the same SMTP session.
2176 54. The store-handling optimization for building very long strings was not
2177 differentiating between the different store pools. I don't think this
2178 actually made any difference in practice, but I've tidied it.
2180 55. While running the routers to verify a sender address, $sender_address
2181 was still set to the sender address. This is wrong, because when routing to
2182 send a bounce to the sender, it would be empty. Therefore, I have changed
2183 it so that, while verifying a sender address, $sender_address is set to <>.
2184 (There is no change to what happens when verifying a recipient address.)
2186 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2187 target A or AAAA records (if not already returned) without resetting the
2188 qualify_single or search_parents options of the DNS resolver. These are
2189 inappropriate in this case because the targets of MX and SRV records must
2190 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2191 target that, when qualified, matched something in the local domain. These
2192 two options are now turned off when doing these lookups.
2194 57. It seems that at least some releases of Reiserfs (which does not have the
2195 concept of a fixed number of inodes) returns zero and not -1 for the
2196 number of available inodes. This interacted badly with check_spool_inodes,
2197 which assumed that -1 was the "no such thing" setting. What I have done is
2198 to check that the total number of inodes is greater than zero before doing
2199 the test of how many are available.
2201 58. When a "warn" ACL statement has a log_message modifier, the message is
2202 remembered, and not repeated. This is to avoid a lot of repetition when a
2203 message has many recipients that cause the same warning to be written.
2204 Howewer, Exim was preserving the list of already written lines for an
2205 entire SMTP session, which doesn't seem right. The memory is now reset if a
2206 new message is started.
2208 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2209 debugging output unless log_rewrite was also set.
2211 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2212 of (int)(handle) when we know that handle contains (void *)(-1).
2214 61. The Exim daemon panic-logs an error return when it closes the incoming
2215 connection. However "connection reset by peer" seems to be common, and
2216 isn't really an error worthy of noting specially, so that particular error
2219 62. When Exim is trying to find all the local interfaces, it used to panic and
2220 die if the ioctl to get the interface flags failed. However, it seems that
2221 on at least one OS (Solaris 9) it is possible to have an interface that is
2222 included in the list of interfaces, but for which you get a failure error
2223 for this call. This happens when the interface is not "plumbed" into a
2224 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2225 failure of the "get flags" call assumes that the interface is down.
2227 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2228 makes life easier for people who are doing arithmetic on fields extracted
2229 from dates, where you often get leading zeros that should not be
2230 interpreted as octal.
2232 64. Added qualify_domain to the redirect router, to override the global
2235 65. If a pathologically long header line contained very many addresses (the
2236 report of this problem mentioned 10 000) and each of them was rewritten,
2237 Exim could use up a very large amount of memory. (It kept on making new
2238 copies of the header line as it rewrote, and never released the old ones.)
2239 At the expense of a bit more processing, the header rewriting function has
2240 been changed so that it no longer eats memory in this way.
2242 66. The generation of the Received: header has been moved from the time that a
2243 message starts to be received, to the time that it finishes. The timestamp
2244 in the Received: header should now be very close to that of the <= log
2245 line. There are two side-effects of this change:
2247 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2248 logged header lines no longer include the local Received: line, because
2249 it has not yet been created. The same applies to a copy of the message
2250 that is returned to a non-SMTP sender when a message is rejected.
2252 (b) When a filter file is tested using -bf, no additional Received: header
2253 is added to the test message. After some thought, I decided that this
2256 This change does not affect the value of $received_for. It is still set
2257 after address rewriting, but before local_scan() is called.
2259 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2261 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2262 gave an unhelpful panic error message, and a defer error. I have managed to
2263 change this behaviour so that it now rejects any supplied certificate,
2264 which seems right, as the list of acceptable certificates is empty.
2266 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2267 gave an unhelpful defer error. I have not managed to make this reject any
2268 supplied certificates, but the error message it gives is "no certificate
2269 supplied", which is not helpful.
2271 70. exigrep's output now also includes lines that are not associated with any
2272 message, but which match the given pattern. Implemented by a patch from
2273 Martin Sluka, which also tidied up the Perl a bit.
2275 71. Recipient callout verification, like sender verification, was using <> in
2276 the MAIL FROM command. This isn't really the right thing, since the actual
2277 sender may affect whether the remote host accepts the recipient or not. I
2278 have changed it to use the actual sender in the callout; this means that
2279 the cache record is now keyed on a recipient/sender pair, not just the
2280 recipient address. There doesn't seem to be a real danger of callout loops,
2281 since a callout by the remote host to check the sender would use <>.
2282 [SEE ABOVE: changed after hitting problems.]
2284 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2285 temporary errors. However, in the case of such a code being given after
2286 the end of a data transmission (i.e. after ".") Exim was failing to write
2287 a retry record for the message. (Yes, there was some broken host that was
2288 actually sending 8xx at this point.)
2290 73. An unknown lookup type in a host list could cause Exim to panic-die when
2291 the list was checked. (An example that provoked this was putting <; in the
2292 middle of a list instead of at the start.) If this happened during a DATA
2293 ACL check, a -D file could be left lying around. This kind of configuration
2294 error no longer causes Exim to die; instead it causes a defer errror. The
2295 incident is still logged to the main and panic logs.
2297 74. Buglet left over from Exim 3 conversion. The message "too many messages
2298 in one connection" was written to the rejectlog but not the mainlog, except
2299 when address rewriting (yes!) was being logged.
2301 75. Added write_rejectlog option.
2303 76. When a system filter was run not as root (that is, when system_filter_user
2304 was set), the values of the $n variables were not being returned to the
2305 main process; thus, they were not subsequently available in the $sn
2308 77. Added +return_path_on_delivery log selector.
2310 78. A connection timeout was being treated differently from recipients deferred
2311 when testing hosts_max_try with a message that was older than the host's
2312 retry timeout. (The host should not be counted, thus allowing all hosts to
2313 be tried at least once before bouncing.) This may have been the cause of an
2314 occasionally reported bug whereby a message would remain on the queue
2315 longer than the retry timeout, but would be bounced if a delivery was
2316 forced. I say "may" because I never totally pinned down the problem;
2317 setting up timeout/retry tests is difficult. See also the next item.
2319 79. The ultimate address timeout was not being applied to errors that involved
2320 a combination of host plus message (for example, a timeout on a MAIL
2321 command). When an address resolved to a number of possible hosts, and they
2322 were not all tried for each delivery (e.g. because of hosts_max_try), a
2323 message could remain on the queue longer than the retry timeout.
2325 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2328 81. Fixed an obscure SMTP outgoing bug which required at least the following
2329 conditions: (a) there was another message waiting for the same server;
2330 (b) the server returned 5xx to all RCPT commands in the first message so
2331 that the message was not completed; (c) the server dropped the connection
2332 or gave a negative response to the RSET that Exim sends to abort the
2333 transaction. The observed case was a dropped connection after DATA that had
2334 been sent in pipelining mode. That is, the server had advertised PIPELINING
2335 but was not implementing it correctly. The effect of the bug was incorrect
2336 behaviour, such as trying another host, and this could lead to a crash.
2342 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2343 and daemon.c were passed as pointers to ints; they should have been
2344 pointers to socklen_t variables (which are typically unsigned ints).
2346 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2349 3. Fixed a really odd bug that affected only the testing scheme; patching a
2350 certain fixed string in the binary changed the value of another string that
2351 happened to be identical to the end of the original first string.
2353 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2354 name", it returns that address as the IP address. On some operating
2355 systems (e.g. Solaris), it also passes back the IP address string as the
2356 "host name". However, on others (e.g. Linux), it passes back an empty
2357 string. Exim wasn't checking for this, and was changing the host name to an
2358 empty string, assuming it had been canonicized.
2360 5. Although rare, it is permitted to have more than one PTR record for a given
2361 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2362 all the names associated with an address, because they do in Solaris.
2363 However, it seems that they do not in Linux for data that comes from the
2364 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2365 I found this out when I moved to a new Linux workstation and tried to run
2366 the Exim test suite.
2368 To get round this problem I have changed the code so that it now does its
2369 own call to the DNS to look up PTR records when searching for a host name.
2370 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2371 addresses that are only in /etc/hosts are still found.
2373 This behaviour is, however, controlled by an option called host_lookup_
2374 order, which defaults to "bydns:byaddr". If people want to use the other
2375 order, or indeed, just use one or the other means of lookup, they can
2376 specify it in this variable.
2378 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2379 some operating systems, this comes back from gethostbyaddr() as an empty
2380 string, and this is what Exim used to test for. However, it seems that in
2381 other systems, "." is yielded. Exim now tests for this case too.
2383 7. The values of check_spool_space and check_log_space are now held internally
2384 as a number of kilobytes instead of an absolute number of bytes. If a
2385 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2386 kilobyte. This means that much larger values can be stored.
2388 8. Exim monitor: an attempt to get the action menu when not actually pointing
2389 at a message produces an empty menu entitled "No message selected". This
2390 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2391 no entries in it ("Shell widget menu has zero width and/or height"). So I
2392 have added a single, blank menu entry in this case.
2394 9. Added ${quote_local_part.
2396 10. MIME decoding is now applied to the contents of Subject: header lines when
2399 11. Now that a reference to $sender_host_address automatically causes a reverse
2400 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2401 host lookup before query-style lookups in lists that might use this
2402 variable. This has therefore been abolished, and the "net-" prefix is no
2403 longer necessary for query-style lookups.
2405 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2406 have been a typo for LFLAGS, so it has been changed.
2408 13. The install script calls Exim with "-C /dev/null" in order to find the
2409 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2410 to be output. Howeve, since Exim outputs its version number before the
2411 error, it didn't break the script. It just looked ugly. I fixed this by
2412 always allowing "-C /dev/null" if the caller is root.
2414 14. Ignore overlarge ACL variable number when reading spool file - insurance
2415 against a later release with more variables having written the file.
2417 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2418 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2419 this, as well as the form without the "IPv6" on the front.
2421 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2422 OS/Makefile-Darwin file.
2424 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2425 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2427 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2428 variable. However, some OS use ints for inodes. I've added cast to long int
2429 to get rid of the compiler warning.
2431 19. I had forgotten to lock out "/../" in configuration file names when
2432 ALT_CONFIG_PREFIX was set.
2434 20. Routers used for verification do not need to specify transports. However,
2435 if such a router generated a host list, and callout was configured, Exim
2436 crashed, because it could not find a port number from the (non-existent)
2437 transport. It now assumes port 25 in this circumstance.
2439 21. Added the -t option to exigrep.
2441 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2442 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2443 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2444 them is set, LOOKUP_LSEARCH is forced.
2446 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2448 24. Added sender and host information to the "rejected by local_scan()" log
2449 line; previously there was no indication of these.
2451 25. Added .include_if_exists.
2453 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2454 Linux. It turns out that not all file systems support this. Apparently some
2455 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2456 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2457 not supported on the file descriptor, is now ignored when Exim is trying to
2458 sync a directory. This applies only to Linux.
2460 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2462 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2463 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2466 29. Added support for maildirsize files from supplied patch (modified a bit).
2468 30. The use of :fail: followed by an empty string could lead Exim to respond to
2469 sender verification failures with (e.g.):
2471 550 Verification failed for <xxx>
2472 550 Sender verify failed
2474 where the first response line was missing the '-' that indicates it is not
2475 the final line of the response.
2477 31. The loop for finding the name of the user that called Exim had a hardwired
2478 limit of 10; it now uses the value of finduser_retries, which is used for
2479 all other user lookups.
2481 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2484 33. Exim was neglecting to zero errno before one call of strtol() when
2485 expanding a string and expecting an integer value. On some systems this
2486 resulted in spurious "integer overflow" errors. Also, it was casting the
2487 result into an int without checking.
2489 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2490 did not work. The code looks as if it has *never* worked, though it appears
2491 to have been documented since at least releast 1.62. I have made it work.
2493 35. The "timeout_DNS" error in retry rules, also documented since at least
2494 1.62, also never worked. As it isn't clear exactly what this means, and
2495 clearly it isn't a major issue, I have abolished the feature by treating it
2496 as "timeout", and writing a warning to the main and panic logs.
2498 36. The display of retry rules for -brt wasn't always showing the error code
2501 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2502 timeout_connect_A, timeout_connect_MX.
2504 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2505 to the empty sender.
2507 39. The daemon was not analysing the content of -oX till after it had closed
2508 stderr and disconnected from the controlling terminal. This meant that any
2509 syntax errors were only noted on the panic log, and the return code from
2510 the command was 0. By re-arranging the code a little, I've made the
2511 decoding happen first, so such errors now appear on stderr, and the return
2512 code is 1. However, the actual setting up of the sockets still happens in
2513 the disconnected process, so errors there are still only recorded on the
2516 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2517 connections (as happens on some IP stacks) was logged at start up time as
2518 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2519 it from "IPv6 and IPv4", which means that two separate sockets are being
2522 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2523 says whether AF_INET or AF_INET6 was passed as an argument.
2525 42. Exiwhat output was messed up when time zones were included in log
2528 43. Exiwhat now gives more information about the daemon's listening ports,
2529 and whether -tls-on-connect was used.
2531 44. The "port" option of the smtp transport is now expanded.
2533 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2534 silently ignored. Now an error message is written to the main and panic
2537 46. There's a new ACL modifier called "logwrite" which writes to a log file
2538 as soon as it is encountered.
2540 47. Added $local_user_uid and $local_user_gid at routing time.
2542 48. Exim crashed when trying to verify a sender address that was being
2545 49. Exim was recognizing only a space character after ".include". It now also
2546 recognizes a tab character.
2548 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2549 extracting the relevant information from the specification. The man page no
2550 longer contains scrambled data for the -d option, and I've added a section
2551 at the front about calling Exim under different names.
2553 51. Added "extra_headers" argument to the "mail" command in filter files.
2555 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2558 53. Installed eximstats 1.29.
2560 54. Added transport_filter_timeout as a generic transport option.
2562 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2563 Cc: header lines. This was required by RFC 822, but it not required by RFC
2566 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2567 incoming messages that did not have them. Now it does so only if the
2568 message originates locally, that is, if there is no associated remote host
2569 address. When Resent- header lines are present, this applies to the Resent-
2570 lines rather than the non-Resent- lines.
2572 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2573 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2575 58. Messages for configuration errors now include the name of the main
2576 configuration file - useful now that there may be more than one file in a
2577 list (.included file names were always shown).
2579 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2580 for those rare installations that do not start the daemon as root or run it
2581 setuid root. I've cut out the call to initgroups() if the daemon is not
2584 60. The Exim user and group can now be bound into the binary as text strings
2585 that are looked up at the start of Exim's processing.
2587 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2589 62. Added $mailstore_basename variable.
2591 63. Installed patch to sieve.c from Michael Haardt.
2593 64. When Exim failed to open the panic log after failing to open the main log,
2594 the original message it was trying to log was written to stderr and debug
2595 output, but if they were not available (the usual case in production), it
2596 was lost. Now it is written to syslog before the two lines that record the
2597 failures to open the logs.
2599 65. Users' Exim filters run in subprocesses under the user's uid. It is
2600 possible for a "deliver" command or an alias in a "personal" command to
2601 provoke an address rewrite. If logging of address rewriting is configured,
2602 this fails because the process is not running as root or exim. There may be
2603 a better way of dealing with this, but for the moment (because 4.30 needs
2604 to be released), I have disabled address rewrite logging when running a
2605 filter in a non-root, non-exim process.
2611 1. The buildconfig auxiliary program wasn't quoting the value set for
2612 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2613 not defined. This bug was masked in 4.22 by the effect that was fixed in
2616 2. Some messages that were rejected after a message id was allocated were
2617 shown as "incomplete" by exigrep. It no longer does this for messages that
2618 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2620 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2621 have allow_domain_literals set, the ID did not get logged in the <= line.
2622 Domain literals are now always recognized in Message-ID: header lines.
2624 4. The first argument for a ${extract expansion item is the key name or field
2625 number. Leading and trailing spaces in this item were not being ignored,
2626 causing some misleading effects.
2628 5. When deliver_drop_privilege was set, single queue runner processes started
2629 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2630 same command in the process it spins off) were not dropping privilege.
2632 6. When the daemon running as "exim" started a queue runner, it always
2633 re-executed Exim in the spun-off process. This is a waste of effort when
2634 deliver_drop_privilege is set. The new process now just calls the
2635 queue-runner function directly.
2641 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2644 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2645 ignored. Though the use of -oP was forcing the writing of a pid file, it
2646 was always written to the default place.
2648 3. If the message "no IP address found for host xxxx" is generated during
2649 incoming verification, it is now followed by identification of the incoming
2650 connection (so you can more easily find what provoked it).
2652 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2654 5. Added some features to "harden" Exim a bit more against certain attacks:
2656 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2657 be put in Local/Makefile. This is like the never_users runtime option,
2658 but it cannot be overridden. The default setting is "root".
2660 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2661 prefix string with which any file named in a -C command line option
2664 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2665 is retained for -C and -D only if the caller of Exim is root. Without
2666 it, the exim user may also use -C and -D and retain privilege.
2668 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2669 command line option is disabled.
2671 6. Macro names set by the -D option must start with an upper case letter, just
2672 like macro names defined in the configuration file.
2674 7. Added "dereference=" facility to LDAP.
2676 8. Two instances of the typo "uknown" in the source files are fixed.
2678 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2679 the Configure-Makefile script screwed up while processing it.
2681 10. Incorporated PCRE 4.4.
2683 11. The SMTP synchronization check was not operating right at the start of an
2684 SMTP session. For example, it could not catch a HELO sent before the client
2685 waited for the greeting. There is now a check for outstanding input at the
2686 point when the greeting is written. Because of the duplex, asynchronous
2687 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2688 way, but not yet received, when the check is performed.
2690 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2691 on TCP/IP sockets, because this apparently causes some broken clients to
2694 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2695 unchanged) from the Cygwin maintainer.
2697 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2698 was supported instead of testing for mbx. Effectively a typo.
2700 15. The spa authenticator server code was not checking that the input it
2701 received was valid base64.
2703 16. The debug output line for the "set" modifier in ACLs was not showing the
2704 name of the variable that was being set.
2706 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2708 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2709 lookup on its own. Something else was needed to trigger the lookup. For
2710 example, a match in host_lookup or the need for a host name in a host list.
2711 Now, if $sender_host_name is referenced and the host name has not yet been
2712 looked up, a lookup is performed. If the lookup fails, the variable remains
2713 empty, and $host_lookup_failed is set to "1".
2715 19. Added "eqi" as a case-independent comparison operator.
2717 20. The saslauthd authentication condition could segfault if neither service
2718 nor realm was specified.
2720 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2721 error message that was logged was misleading, and incoming SMTP
2722 connections were dropped. The message is now more accurate, and temporary
2723 errors are given to SMTP connections.
2725 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2726 (except RSET) with a 421 error, until QUIT is received. However, it was
2727 failing to send a response to QUIT.
2729 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2730 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2731 ACL is now run after the helo_try_verify_hosts code.
2733 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2734 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2735 case-independent, so other case variants are also recognized). Apparently
2736 some systems use these upper case variants.
2738 25. If more than two messages were waiting for the same host, and a transport
2739 filter was specified for the transport, Exim sent two messages over the
2740 same TCP/IP connection, and then failed with "socket operation on non-
2741 socket" when it tried to send the third.
2743 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2745 27. The extern definition of crypt16() in expand.c was not being excluded when
2746 the OS had its own crypt16() function.
2748 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2749 as a preferred synonym for return_size_limit, both as an option and as an
2752 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2754 30. Changed the default configuration ACL to relax the local part checking rule
2755 for addresses that are not in any local domains. For these addresses,
2756 slashes and pipe symbols are allowed within local parts, but the sequence
2757 /../ is explicitly forbidden.
2759 31. SPA server authentication was not clearing the challenge buffer before
2762 32. log_message in a "warn" ACL statement was writing to the reject log as
2763 well as to the main log, which contradicts the documentation and doesn't
2764 seem right (because no rejection is happening). So I have stopped it.
2766 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2767 However, I am unable to do any testing of this.
2769 34. Fixed an infelicity in the appendfile transport. When checking directories
2770 for a mailbox, to see if any needed to be created, it was accidentally
2771 using path names with one or more superfluous leading slashes; tracing
2772 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2774 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2775 contents are added to the log line that is written for every discarded
2776 recipient. (Previously a log_message setting was ignored.)
2778 36. The ${quote: operator now quotes the string if it is empty.
2780 37. The install script runs exim in order to find its version number. If for
2781 some reason other than non-existence or emptiness, which it checks, it
2782 could not run './exim', it was installing it with an empty version number,
2783 i.e. as "exim-". This error state is now caught, and the installation is
2786 38. An argument was missing from the function that creates an error message
2787 when Exim fails to connect to the socket for saslauthd authentication.
2788 This could cause Exim to crash, or give a corrupted message.
2790 39. Added isip, isip4, and isip6 to ${if conditions.
2792 40. The ACL variables $acl_xx are now saved with the message, and can be
2793 accessed later in routers, transports, and filters.
2795 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2796 strings in the file are not string-expanded.
2798 42. If a MAIL command specified a SIZE value that was too large to fit into an
2799 int variable, the check against message_size_limit failed. Such values are
2800 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2801 day this will have to be increased, but I don't think I want to be around
2802 when emails are that large.
2809 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2810 iconv() is not standard in FreeBSD.
2812 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2813 IPv6 enabled. The observed symptom was a segmentation fault on return from
2814 the function os_common_find_running_interfaces() in src/os.c.
2816 3. In the check_special_case() function in daemon.c I had used "errno" as an
2817 argument name, which causes warnings on some systems. This was basically a
2818 typo, since it was named "eno" in the comments!
2820 4. The code that waits for the clock to tick (at a resolution of some fraction
2821 of a second) so as to ensure message-id uniqueness was always waiting for
2822 at least one whole tick, when it could have waited for less. [This is
2823 almost certainly not relevant at current processor speeds, where it is
2824 unlikely to ever wait at all. But we try to future-proof.]
2826 5. The function that sleeps for a time interval that includes fractions of a
2827 second contained a race. It did not block SIGALRM between setting the
2828 timer, and suspending (a couple of lines later). If the interval was short
2829 and the sigsuspend() was delayed until after it had expired, the suspension
2830 never ended. On busy systems this could lead to processes getting stuck for
2833 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2834 process, before it forks any delivery processes. The open lookup caching
2835 mechanism meant that the open file or database connection was passed into
2836 the delivery process. The problem was that delivery processes always tidy
2837 up cached lookup data. This could cause a problem for the next delivery
2838 process started by the queue runner, because the external queue runner
2839 process does not know about the closure. So the next delivery process
2840 still has data in the lookup cache. In the case of a file lookup, there was
2841 no problem because closing a file descriptor in a subprocess doesn't affect
2842 the parent. However, if the lookup was caching a connection to a database,
2843 the connection was closed, and the second delivery process was likely to
2844 see errors such as "PGSQL: query failed: server closed the connection
2845 unexpectedly". The problem has been fixed by closing all cached lookups
2846 in a queue runner before running a delivery process.
2848 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2849 seem to have the "const" qualifier which it has on other OS. I've
2852 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2853 *of the same type* (client or server) with the same public name that an
2854 error should be diagnosed.
2856 9. When Exim looked up a host name for an IP address, but failed to find the
2857 original IP address when looking up the host name (a safety check), it
2858 output the message "<ip address> does not match any IP for NULL", which was
2859 confusing, to say the least. The bug was that the host name should have
2860 appeared instead of "NULL".
2862 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2863 user that is built into the binary, and the -C or -D options is used, root
2864 privilege is dropped before the configuration file is read. In addition,
2865 logging is switched to stderr instead of the normal log files. If the
2866 configuration then re-defines the Exim user, the unprivileged environment
2867 is probably not what is expected, so Exim logs a panic warning message (but
2870 However, if deliver_drop_privilege is set, the unprivileged state may well
2871 be exactly what is intended, so the warning has been cut out in that case,
2872 and Exim is allowed to try to write to its normal log files.
2878 1. smtp_return_error_details was not giving details for temporary sender
2879 or receiver verification errors.
2881 2. Diagnose a configuration error if two authenticators have the same public
2884 3. Exim used not to create the message log file for a message until the first
2885 delivery attempt. This could be confusing when incoming messages were held
2886 for policy or load reasons. The message log file is now created at the time
2887 the message is received, and an initial "Received" line is written to it.
2889 4. The automatically generated man page for command line options had a minor
2890 bug that caused no ill effects; however, a more serious problem was that
2891 the procedure for building the man page automatically didn't always
2892 operate. Consequently, release 4.20 contains an out-of-date version. This
2893 shouldn't happen again.
2895 5. When building Exim with embedded Perl support, the script that builds the
2896 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2897 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2899 6. The freeze_tell option was not being used for messages that were frozen on
2900 arrival, either by an ACL or by local_scan().
2902 7. Added the smtp_incomplete_transaction log selector.
2904 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2905 was accepting AUTH without a new EHLO.
2907 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2908 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2911 10. Logging of TCP/IP connections (when configured) now happens in the main
2912 daemon process instead of the child process, so that the TCP/IP connection
2913 count is more accurate (but it can never be perfect).
2915 11. The use of "drop" in a nested ACL was not being handled correctly in the
2916 outer ACL. Now, if condition failure induced by the nested "drop" causes
2917 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2918 or "require"), the connection is dropped.
2920 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2921 that yield "discard" can now be used with an "accept" or a "discard" verb,
2922 but an error is generated for any others (because I can't see a useful way
2923 to define what should happen).
2925 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2926 are now processed in the same way as lines in the Exim configuration file.
2927 In particular, continuation lines are supported.
2929 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2931 15. Added -ti meaning -t -i.
2933 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2934 domains, and warn by logging if others are found.
2936 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2937 returned from GIFCONF when Exim is trying to find the list of interfaces on
2938 a host. The code in os.c has been modified to copy each ifreq to an aligned
2939 structure in all cases.
2941 Also, in some cases, the returned ifreq's were being copied to a 'struct
2942 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2943 means the last couple of bytes of an IPv6 address could be chopped if the
2944 ifreq contained only a normal sockaddr (14 bytes storage).
2946 18. Named domain lists were not supported in the hosts_treat_as_local option.
2947 An entry such as +xxxx was not recognized, and was treated as a literal
2950 19. Ensure that header lines added by a DATA ACL are included in the reject log
2951 if the ACL subsequently rejects the message.
2953 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2954 MD5 (which is deprecated).
2956 21. When testing a filter file using -bf, Exim was writing a message when it
2957 took the sender from a "From " line in the message, but it was not doing so
2958 when it took $return_path from a Return-Path: header line. It now does.
2960 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2961 with a valid header line field name (a series of printing characters
2962 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2964 23. Changed "disc" in the source to "disk" to conform to the documentation and
2965 the book and for uniformity.
2967 24. Ignore Sendmail's -Ooption=value command line item.
2969 25. When execve() failed while trying to run a command in a pipe transport,
2970 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2971 could be confused with a return value of 69 from the command itself. This
2972 has been changed to 127, the value the shell returns if it is asked to run
2973 a non-existent command. The wording for the related log line suggests a
2974 non-existent command as the problem.
2976 26. If received_header_text expands to an empty string, do not add a Received:
2977 header line to the message. (Well, it adds a token one on the spool, but
2978 marks it "old" so that it doesn't get used or transmitted.)
2980 27. Installed eximstats 1.28 (addition of -nt option).
2982 28. There was no check for failure on the call to getsockname() in the daemon
2983 code. This can fail if there is a shortage of resources on the system, with
2984 ENOMEM, for example. A temporary error is now given on failure.
2986 29. Contrary to the C standard, it seems that in some environments, the
2987 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2988 program. Exim now does this explicitly; it affects the formatting of
2989 timestamps using strftime().
2991 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2992 complaints. I've now initialized all the variables, to avoid this.
2994 32. Header lines added by a system filter were not being "seen" during
2995 transport-time rewrites.
2997 33. The info_callback() function passed to OpenSSL is set up with type void
2998 (*)(SSL *, int, int), as described somewhere. However, when calling the
2999 function (actually a macro) that sets it up, the type void(*)() is
3000 expected. I've put in a cast to prevent warnings from picky compilers.
3002 34. If a DNS black list lookup found a CNAME record, but there were no A
3003 records associated with the domain it pointed at, Exim crashed.
3005 35. If a DNS black list lookup returned more than one A record, Exim ignored
3006 all but the first. It now scans all returned addresses if a particular IP
3007 value is being sought. In this situation, the contents of the
3008 $dnslist_value variable are a list of all the addresses, separated by a
3011 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3012 to accept a host name and all its aliases if the forward lookup for any of
3013 them yielded the IP address of the incoming connection. Now it accepts only
3014 those names whose forward lookup yields the correct IP address. Any other
3015 names are discarded. This closes a loophole whereby a rogue DNS
3016 administrator could create reverse DNS records to break through a
3017 wildcarded host restriction in an ACL.
3019 37. If a user filter or a system filter that ran in a subprocess used any of
3020 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3021 the wrong values were passed to the pipe command ($thisaddress had the
3022 value of $0, $0 had the value of $1, etc). This bug was introduced by
3023 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3025 38. Improved the line breaking for long SMTP error messages from ACLs.
3026 Previously, if there was no break point between 40 and 75 characters, Exim
3027 left the rest of the message alone. Two changes have been made: (a) I've
3028 reduced the minimum length to 35 characters; (b) if it can't find a break
3029 point between 35 and 75 characters, it looks ahead and uses the first one
3030 that it finds. This may give the occasional overlong line, but at least the
3031 remaining text gets split now.
3033 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3034 file descriptors might be low, and that setting 1000 would always raise it.
3035 It turns out that in some environments, the limit is already over 1000 and
3036 that lowering it causes trouble. So now Exim takes care not to decrease it.
3038 40. When delivering a message, the value of $return_path is set to $sender_
3039 address at the start of routing (routers may change the value). By an
3040 oversight, this default was not being set up when an address was tested by
3041 -bt or -bv, which affected the outcome if any router or filter referred to
3044 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3045 log only when "message" or "log_message" are set. However, if one of the
3046 conditions was an address verification, or a call to a nested ACL, the
3047 messages generated by the underlying test were being passed through. This
3048 no longer happens. The underlying message is available in $acl_verify_
3049 message for both "message" and "log_message" expansions, so it can be
3050 passed through if needed.
3052 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3053 new expansion $bh_ to give the encoded byte string without charset
3054 translation. Translation happens only if iconv() is available; HAVE_ICONV
3055 indicates this at build time. HEADERS_CHARSET gives the charset to
3056 translate to; headers_charset can change it in the configuration, and
3057 "headers charset" can change it in an individual filter file.
3059 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3060 that creates RFC 2047 encoded "words" labels them as that charset instead
3061 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3062 expansion operator; (ii) when Exim creates a From: line for a local
3063 message; (iii) when a header line is rewritten to include a "phrase" part.
3065 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3066 buggy, causing it to skip the first lines of messages whose message ID
3067 ended in 'D'. This would not have bitten before Exim release 4.14, because
3068 message IDs were unlikely to end in 'D' before then. The effect was to have
3069 incorrect size information for certain domains.
3071 45. #include "config.h" was missing at the start of the crypt16.c module. This
3072 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3075 46. If there was a timeout during a "random" callout check, Exim treated it as
3076 a failure of the random address, and carried on sending RSET and the real
3077 address. If the delay was just some slowness somewhere, the response to the
3078 original RCPT would be taken as a response to RSET and so on, causing
3079 mayhem of various kinds.
3081 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3082 when I implemented it. It didn't allow for the fact that some option values
3083 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3084 the right test for positive values.
3086 48. Domain names in DNS records are case-independent. Exim always looks them up
3087 in lower case. Some resolvers return domain names in exactly the case they
3088 appear in the zone file, that is, they may contain uppercase letters. Not
3089 all resolvers do this - some return always lower case. Exim was treating a
3090 change of case by a resolver as a change of domain, similar to a widening
3091 of a domain abbreviation. This triggered its re-routing code and so it was
3092 trying to route what was effectively the same domain again. This normally
3093 caused routing to fail (because the router wouldn't handle the domain
3094 twice). Now Exim checks for this case specially, and just changes the
3095 casing of the domain that it ultimately uses when it transmits the message
3098 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3101 50. If a filter generated a file delivery with a non-absolute name (possible if
3102 no home directory exists for the router), the forbid_file option was not
3105 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3106 the existing equality matching.
3108 52. Exim was using ints instead of ino_t variables in some places where it was
3109 dealing with inode numbers.
3111 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3112 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3113 called TMPDIR, and if it finds it is different, it changes its value.
3115 54. The smtp_printf() function is now made available to local_scan() so
3116 additional output lines can be written before returning. There is also an
3117 smtp_fflush() function to enable the detection of a dropped connection.
3118 The variables smtp_input and smtp_batched_input are exported to
3121 55. Changed the default runtime configuration: the message "Unknown user"
3122 has been removed from the ACL, and instead placed on the localuser router,
3123 using the cannot_route_message feature. This means that any verification
3124 failures that generate their own messages won't get overridden. Similarly,
3125 the "Unrouteable address" message that was in the ACL for unverifiable
3126 relay addresses has also been removed.
3128 56. Added hosts_avoid_esmtp to the smtp transport.
3130 57. The exicyclog script was not checking for the esoteric option
3131 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3132 will work only if exicyclog is run under the appropriate euid.
3134 58. Following a discussion on the list, the rules by which Exim recognises line
3135 endings on incoming messages have been changed. The -dropcr and drop_cr
3136 options are now no-ops, retained only for backwards compatibility. The
3137 following line terminators are recognized: LF CRLF CR. However, special
3138 processing applies to CR:
3140 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3141 nor a local message in the state where . is a terminator.
3143 (ii) If a bare CR is encountered in a header line, an extra space is added
3144 after the line terminator so as not to end the header. The reasoning
3145 behind this is that bare CRs in header lines are most likely either
3146 to be mistakes, or people trying to play silly games.
3148 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3149 was being incorrectly given as 18 bytes larger than it should have been.
3150 This is a VOB (very old bug).
3152 60. This may never have affected anything current, but just in case it has:
3153 When the local host is found other than at the start of a list of hosts,
3154 the local host, those with the same MX, and any that follow, are discarded.
3155 When the list in question was part of a longer list of hosts, the following
3156 hosts (not currently being processed) were also being discarded. This no
3157 longer happens. I'm not sure if this situation could ever has previously
3160 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3163 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3164 header. This is something that is recommended in a new Internet Draft, and
3165 is something that is documented as being done by Sendmail. There are two
3166 possible values. For messages generated by the autoreply transport, Exim
3169 Auto-Submitted: auto-replied
3171 whereas for all other generated messages (e.g. bounces) it adds
3173 Auto-Submitted: auto-generated
3175 63. The "personal" condition in filters now includes a test for the
3176 Auto-Submitted: header. If it contains the string "auto-" the message it
3177 not considered personal.
3179 64. Added rcpt_include_affixes as a generic transport option.
3181 65. Added queue_only_override (default true).
3183 66. Added the syslog_duplication option.
3185 67. If what should have been the first header line of a message consisted of
3186 a space followed by a colon, Exim was mis-interpreting it as a header line.
3187 It isn't of course - it is syntactically invalid and should therefore be
3188 treated as the start of the message body. The misbehaviour could have
3189 caused a number of strange effects, including loss of data in subsequent
3190 header lines, and spool format errors.
3192 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3193 client host had authenticated. This control can now be exercised by an ACL
3194 for more flexibility.
3196 69. By default, callouts do not happen when testing with -bh. There is now a
3197 variant, -bhc, which does actually run the callout code, including
3198 consulting and updating the callout cache.
3200 70. Added support for saslauthd authentication, courtesy of Alexander
3203 71. If statvfs() failed on the spool or log directories while checking their
3204 size for availability, Exim confusingly gave the error "space shortage".
3205 Furthermore, in debugging mode it crashed with a floating point exception.
3206 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3207 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3208 serious problem, Exim now writes to the main and panic logs when this
3209 happens, with details of the failure. It then refuses to accept the
3210 incoming message, giving the message "spool directory problem" or "log
3211 directory problem" with a 421 code for SMTP messages.
3213 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3214 0, 1, and 2 exist, because some OS complain for execs without them (see
3215 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3216 descriptors. However, the code omitted to check that the open succeeded,
3217 causing mysterious errors if for some reason the permissions on /dev/null
3218 got screwed. Now Exim writes a message to the main and panic logs, and
3219 bombs out if it can't open /dev/null.
3221 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3222 interact so that it is all more flexible. It is supposed to remain
3223 backwards compatible. Also added extra_local_interfaces.
3225 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3226 to bomb out with an assertion failure - to the client this appears as a
3227 connection drop. This problem occurs in the part of the code that was taken
3228 from the Samba project. Fortunately, the assertion is in a very simple
3229 function, so I have fixed this by reproducing the function inline in the
3230 one place where it is called, and arranging for authentication to fail
3231 instead of killing the process with assert().
3233 75. The SPA client code was not working when the server requested OEM rather
3234 than Unicode encoding.
3236 76. Added code to make require_files with a specific uid setting more usable in
3237 the case where statting the file as root fails - usually a non-root-mounted
3238 NFS file system. When this happens and the failure is EACCES, Exim now
3239 forks a subprocess and does the per-uid checking as the relevant uid.
3241 77. Added process_log_path.
3243 78. If log_file_path was not explicitly set, a setting of check_log_space or
3244 check_log_inodes was ignored.
3246 79. If a space check for the spool or log partitions fails, the incident is now
3247 logged. Of course, in the latter case the data may get lost...
3249 80. Added the %p formatting code to string_format() so that it can be used to
3250 print addresses in debug_print(). Adjusted all the address printing in the
3251 debugging in store.c to use %p rather than %d.
3253 81. There was a concern that a line of code in smtp_in.c could overflow a
3254 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3255 initially expressed, the concern was not well-founded, because trailing
3256 spaces are removed early. However, if the trailing spaces were followed by
3257 a NULL, they did not get removed, so the overflow was possible. Two fixes
3260 (a) I re-wrote the offending code in a cleaner fashion.
3261 (b) If an incoming SMTP command contains a NULL character, it is rejected
3264 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3265 runs initgroups(), so that if the Exim user is in any additional groups,
3266 they will be used during message reception.
3272 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob