1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.204 2005/08/09 13:31:52 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".
90 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
92 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
93 command itself ended in error, the underlying message about the transport
94 filter timeout was being overwritten with the pipe command error. Now the
95 underlying error message should be appended to the second error message.
97 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
99 PH/02 Reduce the amount of output that "make" produces by default. Full output
100 can still be requested.
102 PH/03 The warning log line about a condition test deferring for a "warn" verb
103 was being output only once per connection, rather than after each
104 occurrence (because it was using the same function as for successful
105 "warn" verbs). This seems wrong, so I have changed it.
107 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
108 it should not have, which might have caused a crash in the right
109 circumstances, but probably never did.
111 PH/04 Installed a modified version of Tony Finch's patch to make submission
112 mode fix the return path as well as the Sender: header line, and to
113 add a /name= option so that you can make the user's friendly name appear
116 TF/03 Added the control = fakedefer ACL modifier.
118 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
119 Mark Lowes for thorough testing.
121 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
123 TK/03 Merged latest SRS patch from Miles Wilton.
125 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
126 with the definition in sysexits.h (which is #included earlier).
127 Fortunately, Exim does not actually use EX_OK. The code used to try to
128 preserve the sysexits.h value, by assumimg that macro definitions were
129 scanned for macro replacements. I have been disabused of this notion,
130 so now the code just undefines EX_OK before #including unistd.h.
132 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
133 in the smtp transport. When a block could not be written in a single
134 write() function, the timeout was being re-applied to each part-write.
135 This seems wrong - if the receiver was accepting one byte at a time it
136 would take for ever. The timeout is now adjusted when this happens. It
137 doesn't have to be particularly precise.
139 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
140 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
142 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
143 without a domain if the check to <postmaster@domain> fails.
145 SC/01 Eximstats: added -xls and the ability to specify output files
146 (patch written by Frank Heydlauf).
148 SC/02 Eximstats: use FileHandles for outputing results.
150 SC/03 Eximstats: allow any combination of xls, txt, and html output.
152 SC/04 Eximstats: fixed display of large numbers with -nvr option
154 SC/05 Eximstats: fixed merging of reports with empty tables.
156 SC/06 Eximstats: added the -include_original_destination flag
158 SC/07 Eximstats: removed tabs and trailing whitespace.
160 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
162 TK/06 MBOX spool code: Add real "From " MBOX separator line
163 so the .eml file is really in mbox format (even though
164 most programs do not really care). Patch from Alex Miller.
166 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
167 The latter is generated from $received_to and is only set if the
168 message has one envelope recipient. SA can use these headers,
169 obviously out-of-the-box. Patch from Alex Miller.
171 PH/08 The ${def test on a variable was returning false if the variable's
172 value was "0", contrary to what the specification has always said!
173 The result should be true unless the variable is empty.
175 PH/09 The syntax error of a character other than { following "${if
176 def:variable_name" (after optional whitespace) was not being diagnosed.
177 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
178 accidental colon was present, for example, could give incorrect results.
180 PH/10 Tidied the code in a number of places where the st_size field of a stat()
181 result is used (not including appendfile, where other changes are about
184 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
185 This involved changing a lot of size variables from int to off_t. It
186 should work with maildirs and everything.
188 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
189 spamd dying while we are connected to it.
191 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
194 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
195 function for BDB 4.3.
197 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
199 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
200 the sources for PCRE have been re-arranged and more files are now
203 PH/15 The code I had for printing potentially long long variables in PH/11
204 above was not the best (it lost precision). The length of off_t variables
205 is now inspected at build time, and an appropriate printing format (%ld
206 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
207 to be "long long int" or "long int". This is needed for the internal
208 formatting function string_vformat().
210 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
211 the configuration file to be ":syslog", then the script "guesses" where
212 the logs files are, rather than using the compiled in default. In our
213 case the guess is not the same as the compiled default, so the script
214 suddenly stopped working when I started to use syslog. The patch checks
215 to see if log_file_path is "". If so, it attempts to read it from exim
216 with no configuration file to get the compiled in version, before it
217 falls back to the previous guessing code."
219 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
220 implementing BATV in an Exim configuration. See NewStuff for the gory
223 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
224 Makefile that are specific to HP-UX.
226 PH/18 If the "use_postmaster" option was set for a recipient callout together
227 with the "random" option, the postmaster address was used as the MAIL
228 FROM address for the random test, but not for the subsequent recipient
229 test. It is now used for both.
231 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
232 patch removes a few documentation additions to RFC 3028, because the
233 latest draft now contains them. It adds the new en;ascii-case comparator
234 and a new error check for 8bit text in MIME parts. Comparator and
235 require names are now matched exactly. I enabled the subaddress
236 extension, but it is not well tested yet (read: it works for me)."
238 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
239 rework some of the code of TK/09 above to avoid the hardwired use of
240 "%lld" and "long long". Replaced the call to snprintf() with a call to
243 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
244 records point to non-existent hosts", "retry timeout exceeded", and
245 "retry time not reached for any host after a long failure period".
247 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
248 experimental DomainKeys support:
250 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
251 (2) On an error such as an illegally used "control", the wrong name for
252 the control was given.
254 These problems did NOT occur unless DomainKeys support was compiled.
256 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
258 PH/24 Added ${if match_ip condition.
260 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
261 databases so that it will be absolutely obvious if a crash occurs in the
262 DB library. This is a regular occurrence (often caused by mis-matched
265 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
266 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
267 picked up on a user's system that detects such things. There doesn't seem
268 to be a gcc warning option for this - only an attribute that has to be
269 put on the function's prototype. It seems that in Fedora Core 4 they have
270 set this on a number of new functions. No doubt there will be more in due
273 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
274 specify a transport. However, if an address that was verified by such a
275 router was the subject of a callout, Exim crashed because it tried to
276 read the rcpt_include_affixes from the non-existent transport. Now it
277 just assumes that the setting of that option is false. This bug was
278 introduced by 4.51/PH/31.
280 PH/28 Changed -d+all to exclude +memory, because that information is very
281 rarely of interest, but it makes the output a lot bigger. People tend to
282 do -d+all out of habit.
284 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
285 code in os-type was giving problems when libc.so lives in lib64, like on
288 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
289 aren't the modern standard, and it seems that some systems' include files
290 don't always have them. Exim was already checking for some of the newer
291 ones like T_AAAA, and defining it itself. I've added checks for all the
292 record types that Exim uses.
294 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
295 not automatically generating a new one, as it is supposed to. This
296 prevented TLS from working. If the file did exist, but contained invalid
297 data, a new version was generated, as expected. It was only the case of a
298 non-existent file that was broken.
300 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
301 with a change in libdomainkeys > 0.64.
303 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
304 from DNS. If the selector record carries the flag, it now has
305 precedence over the domain-wide flag.
307 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
309 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
310 the use of a port name that isn't defined in /etc/services) occurred, the
311 message was deferred as in a normal delivery, and thus remained on the
312 spool, instead of being failed because of the mua_wrapper setting. This
313 is now fixed, and I tidied up some of the mua_wrapper messages at the
316 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
317 the messages in a hash of arrays rather than using individual hashes.
318 This is a bit cleaner and results in dramatic memory savings, albeit
319 at a slight CPU cost.
321 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
322 as requested by Marc Sherman.
324 SC/10 Eximstats: added histograms for user specified patterns as requested
327 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
329 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
330 fopen() in the content-scanning modules that did not already have it.
336 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
337 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
339 TK/02 Fix ACL "control" statement not being available in MIME ACL.
341 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
343 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
344 to test Sieve filters that use "vacation".
346 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
347 that changes the way the GnuTLS parameters are stored in the cache file.
348 The new format can be generated externally. For backward compatibility,
349 if the data in the cache doesn't make sense, Exim assumes it has read an
350 old-format file, and it generates new data and writes a new file. This
351 means that you can't go back to an older release without removing the
354 PH/03 A redirect router that has both "unseen" and "one_time" set does not
355 work if there are any delivery delays because "one_time" forces the
356 parent to be marked "delivered", so its unseen clone is never tried
357 again. For this reason, Exim now forbids the simultaneous setting of
360 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
361 redirected to themselves ("homonym" addresses). Read the long ChangeLog
362 entry if you want to know the details. The fix, however, neglected to
363 consider the case when local delivery batching is involved. The test for
364 "previously delivered" was not happening when checking to see if an
365 address could be batched with a previous (undelivered) one; under
366 certain circumstances this could lead to multiple deliveries to the same
369 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
370 in its include files, and this causes problems building Exim.
372 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
373 header_syntax) but Exim was just ignoring anything given after a slash.
374 In particular, this caused confusion with an attempt to use "verify =
375 reverse_host_lookup/defer_ok". An error is now given when options are
376 supplied for verify items that do not have them. (Maybe reverse_host_
377 lookup should have a defer_ok option, but that's a different point.)
379 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
380 defined by RFC 821) to 2048, because there were problems with some AUTH
381 commands, and RFC 1869 says the size should be increased for extended
382 SMTP commands that take arguments.
384 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
387 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
388 "unknown" error; now it says that the functionality isn't in the binary.
390 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
391 an address' error message when a string expansion fails (syntax or
392 whatever). Otherwise the password may appear in the log. Following change
393 PH/42 below, there is no longer a chance of it appearing in a bounce
396 PH/11 Installed exipick version 20050225.0 from John Jetmore.
398 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
399 of its addresses was ever tried. (Bugzilla bug #2.)
401 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
402 the result incorrectly in the debug output. (It correctly added a newline
403 to what was transported.)
405 TF/01 Added $received_time.
407 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
408 commented out examples of how to interface to a virus scanner and to
409 SpamAssassin. Also added commented examples of av_scanner and
410 spamd_address settings.
412 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
413 and controls are allowed in which ACLs. There were a couple of minor
414 errors. Some of the entries in the conditions table (which is a table of
415 where they are NOT allowed) were getting very unwieldy; rewrote them as a
416 negation of where the condition IS allowed.
418 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
420 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
421 header file does not have a version number, so I've had to invent a new
422 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
423 API. The code is untested by me (my Linux distribution still has 0.3.2 of
424 radiusclient), but it was contributed by a Radius user.
426 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
427 files or directories, for OpenSSL.
429 PH/19 When an Exim process that is running as root has to create an Exim log
430 file, it does so in a subprocess that runs as exim:exim so as to get the
431 ownership right at creation (otherwise, other Exim processes might see
432 the file with the wrong ownership). There was no test for failure of this
433 fork() call, which would lead to the process getting stuck as it waited
434 for a non-existent subprocess. Forks do occasionally fail when resources
435 run out. I reviewed all the other calls to fork(); they all seem to check
438 PH/20 When checking for unexpected SMTP input at connect time (before writing
439 the banner), Exim was not dealing correctly with a non-positive return
440 from the read() function. If the client had disconnected by this time,
441 the result was a log entry for a synchronization error with an empty
442 string after "input=" when read() returned zero. If read() returned -1
443 (an event I could not check), uninitialized data bytes were printed.
444 There were reports of junk text (parts of files, etc) appearing after
447 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
449 PH/22 Added support for macro redefinition, and (re)definition in between
450 driver and ACL definitions.
452 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
453 forgetting to use the resulting value; it was using the unexpanded value.
455 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
456 hadn't been configured. The fix is from Juergen Kreileder, who
457 understands it better than I do:
459 "Here's what I see happening with three configured cyrus_sasl
460 authenticators configured (plain, login, cram-md5):
462 On startup auth_cyrus_sasl_init() gets called for each of these.
463 This means three calls to sasl_listmech() without a specified mech_list.
464 => SASL tests which mechs of all available mechs actually work
465 => three warnings about OTP not working
466 => the returned list contains: plain, login, cram-md5, digest-md5, ...
468 With the patch, sasl_listmech() also gets called three times. But now
469 SASL's mech_list option is set to the server_mech specified in the the
470 authenticator. Or in other words, the answer from sasl_listmech()
471 gets limited to just the mech you're testing for (which is different
473 => the return list contains just 'plain' or 'login', 'cram-md5' or
474 nothing depending on the value of ob->server_mech.
476 I've just tested the patch: Authentication still works fine,
477 unavailable mechs specified in the exim configuration are still
478 caught, and the auth.log warnings about OTP are gone."
480 PH/25 When debugging is enabled, the contents of the command line are added
481 to the debugging output, even when log_selector=+arguments is not
484 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
485 answer is "GNU", and only if the return is "GNU/something" is the answer
488 PH/27 $acl_verify_message is now set immediately after the failure of a
489 verification in an ACL, and so is available in subsequent modifiers. In
490 particular, the message can be preserved by coding like this:
492 warn !verify = sender
493 set acl_m0 = $acl_verify_message
495 Previously, $acl_verify_message was set only while expanding "message"
496 and "log_message" when a very denied access.
498 PH/28 Modified OS/os.c-Linux with
500 -#ifndef OS_LOAD_AVERAGE
501 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
503 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
504 nomenclature these days.)
506 PH/29 Installed patch from the Sieve maintainer that adds the options
507 sieve_useraddress and sieve_subaddress to the redirect router.
509 PH/30 In these circumstances:
510 . Two addresses routed to the same list of hosts;
511 . First host does not offer TLS;
512 . First host accepts first address;
513 . First host gives temporary error to second address;
514 . Second host offers TLS and a TLS session is established;
515 . Second host accepts second address.
516 Exim incorrectly logged both deliveries with the TLS parameters (cipher
517 and peerdn, if requested) that were in fact used only for the second
520 PH/31 When doing a callout as part of verifying an address, Exim was not paying
521 attention to any local part prefix or suffix that was matched by the
522 router that accepted the address. It now behaves in the same way as it
523 does for delivery: the affixes are removed from the local part unless
524 rcpt_include_affixes is set on the transport.
526 PH/32 Add the sender address, as F=<...>, to the log line when logging a
527 timeout during the DATA phase of an incoming message.
529 PH/33 Sieve envelope tests were broken for match types other than :is. I have
530 applied a patch sanctioned by the Sieve maintainer.
532 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
533 the uid or gid is negative. A case of a negative gid caused this to be
534 noticed. The fix allows for either to be negative.
536 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
537 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
540 PH/36 The change PH/12 above was broken. Fixed it.
542 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
543 the grounds that routing the same address twice would always produce the
544 same answer. This might have been true once, but it is certainly no
545 longer true now. Routing a child address may depend on the previous
546 routing that produced that child. Some complicated redirection strategies
547 went wrong when messages had multiple recipients, and made Exim's
548 behaviour dependent on the order in which the addresses were given.
550 I have moved the duplicate checking until after the routing is complete.
551 Exim scans the addresses that are assigned to local and remote
552 transports, and removes any duplicates. This means that more work will be
553 done, as duplicates will always all be routed, but duplicates are
554 presumably rare, so I don't expect this is of any significance.
556 For deliveries to pipes, files, and autoreplies, the duplicate checking
557 still happens during the routing process, since they are not going to be
560 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
561 It corrects a timeout issue with spamd. This is Ian's comment: "The
562 background is that sometimes spamd either never reads data from a
563 connection it has accepted, or it never writes response data. The exiscan
564 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
565 blindly assumes that writes won't block so it may never time out."
567 PH/39 Allow G after quota size as well as K and M.
569 PH/40 The value set for $authenticated_id in an authenticator may not contain
570 binary zeroes or newlines because the value is written to log lines and
571 to spool files. There was no check on this. Now the value is run through
572 the string_printing() function so that such characters are converted to
573 printable escape sequences.
575 PH/41 $message_linecount is a new variable that contains the total number of
576 lines in the message. Compare $body_linecount, which is the count for the
579 PH/42 Exim no longer gives details of delivery errors for specific addresses in
580 bounce and delay warning messages, except in certain special cases, which
583 (a) An SMTP error message from a remote host;
584 (b) A message specified in a :fail: redirection;
585 (c) A message specified in a "fail" command in a system filter;
586 (d) A message specified in a FAIL return from the queryprogram router;
587 (e) A message specified by the cannot_route_message router option.
589 In these cases only, Exim does include the error details in bounce and
590 warning messages. There are also a few cases where bland messages such
591 as "unrouteable address" or "local delivery error" are given.
593 PH/43 $value is now also set for the "else" part of a ${run expansion.
595 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
596 being worked on, but at least Exim now implements the latest version to
599 PH/45 In a pipe transport, although a timeout while waiting for the pipe
600 process to complete was treated as a delivery failure, a timeout while
601 writing the message to the pipe was logged, but erroneously treated as a
602 successful delivery. Such timeouts include transport filter timeouts. For
603 consistency with the overall process timeout, these timeouts are now
604 treated as errors, giving rise to delivery failures by default. However,
605 there is now a new Boolean option for the pipe transport called
606 timeout_defer, which, if set TRUE, converts the failures into defers for
607 both kinds of timeout. A transport filter timeout is now identified in
610 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
611 systems where "make" and "gmake" are different, calling "gmake" at top
612 level broke things. I've arranged for the value of $(MAKE) to be passed
613 from the Makefile to this script so that it can call the same version of
617 A note about Exim versions 4.44 and 4.50
618 ----------------------------------------
620 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
621 changes of various kinds. As a consequence, a big documentation update was
622 needed. This delayed the release for rather longer than seemed good, especially
623 in the light of a couple of (minor) security issues. Therefore, the changes
624 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
625 release. So 4.44 and 4.50 are in effect two different branches that both start
628 I have left the 4.50 change log unchanged; it contains all the changes since
629 4.43. The change log for 4.44 is below; many of its items are identical to
630 those for 4.50. This seems to be the most sensible way to preserve the
631 historical information.
637 1. Minor wording change to the doc/README.SIEVE file.
639 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
640 computation of the current number of files was incorrect.
642 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
643 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
646 4. Give more explanation in the error message when the command for a transport
647 filter fails to execute.
649 5. There are several places where Exim runs a non-Exim command in a
650 subprocess. The SIGUSR1 signal should be disabled for these processes. This
651 was being done only for the command run by the queryprogram router. It is
652 now done for all such subprocesses. The other cases are: ${run, transport
653 filters, and the commands run by the lmtp and pipe transports.
655 6. Added CONFIGURE_GROUP build-time option.
657 7. Some older OS have a limit of 256 on the maximum number of file
658 descriptors. Exim was using setrlimit() to set 1000 as a large value
659 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
660 systems. I've change it so that if it can't get 1000, it tries for 256.
662 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
663 was an oversight, and furthermore, ever since the addition of extra
664 controls (e.g. 4.43/32), the checks on when to allow different forms of
665 "control" were broken. There should now be diagnostics for all cases when a
666 control that does not make sense is encountered.
668 9. Added the /retain_sender option to "control=submission".
670 10. $recipients is now available in the predata ACL (oversight).
672 11. Tidy the search cache before the fork to do a delivery from a message
673 received from the command line. Otherwise the child will trigger a lookup
674 failure and thereby defer the delivery if it tries to use (for example) a
675 cached ldap connection that the parent has called unbind on.
677 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
678 of $address_data from the recipient verification was clobbered by the
681 13. The value of address_data from a sender verification is now available in
682 $sender_address_data in subsequent conditions in the ACL statement.
684 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
686 15. Added a new option "connect=<time>" to callout options, to set a different
689 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
690 was its contents. (It was OK if the option was not defined at all.)
692 17. A "Completed" log line is now written for messages that are removed from
693 the spool by the -Mrm option.
695 18. New variables $sender_verify_failure and $recipient_verify_failure contain
696 information about exactly what failed.
698 19. Added -dd to debug only the daemon process.
700 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
701 handles timeouts, both on the server side and network timeouts. Renamed the
702 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
704 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
705 It is now set to "smtps".
707 22. $host_address is now set to the target address during the checking of
710 23. When checking ignore_target_hosts for an ipliteral router, no host name was
711 being passed; this would have caused $sender_host_name to have been used if
712 matching the list had actually called for a host name (not very likely,
713 since this list is usually IP addresses). A host name is now passed as
716 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
717 code that specifies a non-restarting handler (typically sigaction() in
718 modern systems) in an attempt to fix a rare and obscure crash bug.
720 25. Narrowed the window for a race in the daemon that could cause it to ignore
721 SIGCHLD signals. This is not a major problem, because they are used only to
722 wake it up if nothing else does.
724 26. A malformed maildirsize file could cause Exim to calculate negative values
725 for the mailbox size or file count. Odd effects could occur as a result.
726 The maildirsize information is now recalculated if the size or filecount
729 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
730 support for a long time. Removed HAVE_SYS_VFS_H.
732 28. Installed the latest version of exipick from John Jetmore.
734 29. In an address list, if the pattern was not a regular expression, an empty
735 subject address (from a bounce message) matched only if the pattern was an
736 empty string. Non-empty patterns were not even tested. This was the wrong
737 because it is perfectly reasonable to use an empty address as part of a
738 database query. An empty address is now tested by patterns that are
739 lookups. However, all the other forms of pattern expect the subject to
740 contain a local part and a domain, and therefore, for them, an empty
741 address still always fails if the pattern is not itself empty.
743 30. Exim went into a mad DNS loop when attempting to do a callout where the
744 host was specified on an smtp transport, and looking it up yielded more
747 31. Re-factored the code for checking spool and log partition space into a
748 function that finds that data and another that does the check. The former
749 is then used to implement four new variables: $spool_space, $log_space,
750 $spool_inodes, and $log_inodes.
752 32. The RFC2047 encoding function was originally intended for short strings
753 such as real names; it was not keeping to the 75-character limit for
754 encoded words that the RFC imposes. It now respects the limit, and
755 generates multiple encoded words if necessary. To be on the safe side, I
756 have increased the buffer size for the ${rfc2047: expansion operator from
759 33. It is now permitted to omit both strings after an "if" condition; if the
760 condition is true, the result is "true". As before, when the second string
761 is omitted, a false condition yields an empty string. This makes it less
762 cumbersome to write custom ACL and router conditions.
764 34. Failure to deliver a bounce message always caused it to be frozen, even if
765 there was an errors_to setting on the router. The errors_to setting is now
768 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
769 canonical form (fully expanded) before being placed in
770 $sender_host_address.
772 36. The table in the code that translates DNS record types into text (T_A to
773 "A" for instance) was missing entries for NS and CNAME. It is just possible
774 that this could have caused confusion if both these types were looked up
775 for the same domain, because the text type is used as part of Exim's
776 per-process caching. But the chance of anyone hitting this buglet seems
779 37. The dnsdb lookup has been extended in a number of ways.
781 (1) There is a new type, "zns", which walks up the domain tree until it
782 finds some nameserver records. It should be used with care.
784 (2) There is a new type, "mxh", which is like "mx" except that it returns
785 just the host names, not the priorities.
787 (3) It is now possible to give a list of domains (or IP addresses) to be
788 looked up. The behaviour when one of the lookups defers can be
789 controlled by a keyword.
791 (4) It is now possible to specify the separator character for use when
792 multiple records are returned.
794 38. The dnslists ACL condition has been extended: it is now possible to supply
795 a list of IP addresses and/or domains to be looked up in a particular DNS
798 39. Added log_selector=+queue_time_overall.
800 40. When running the queue in the test harness, wait just a tad after forking a
801 delivery process, to get repeatability of debugging output.
803 41. Include certificate and key file names in error message when GnuTLS fails
804 to set them up, because the GnuTLS error message doesn't include the name
805 of the failing file when there is a problem reading it.
807 42. Allow both -bf and -bF in the same test run.
809 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
811 44. The "Exiscan patch" is now merged into the mainline Exim source.
813 45. Sometimes the final signoff response after QUIT could fail to get
814 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
815 < 0 before doing a fflush(). This bug looks as though it goes back to the
816 introduction of TLS in release 3.20, but "sometimes" must have been rare
817 because the tests only now provoked it.
819 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
820 (this can affect the format of dates).
822 47. exim_tidydb, when checking for the continued existence of a message for
823 which it has found a message-specific retry record, was not finding
824 messages that were in split spool directories. Consequently, it was
825 deleting retry records that should have stayed in existence.
827 48. Steve fixed some bugs in eximstats.
829 49. The SPA authentication driver was not abandoning authentication and moving
830 on to the next authenticator when an expansion was forced to fail,
831 contradicting the general specification for all authenticators. Instead it
832 was generating a temporary error. It now behaves as specified.
834 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
835 (the order specifies the preference for clients). The order is now AES256,
836 AES128, 3DES, ARCFOUR128.
838 51. Small patch to Sieve code - explicitly set From: when generating an
841 52. Exim crashed if a remote delivery caused a very long error message to be
842 recorded - for instance if somebody sent an entire SpamAssassin report back
843 as a large number of 550 error lines. This bug was coincidentally fixed by
844 increasing the size of one of Exim's internal buffers (big_buffer) that
845 happened as part of the Exiscan merge. However, to be on the safe side, I
846 have made the code more robust (and fixed the comments that describe what
849 53. Now that there can be additional text after "Completed" in log lines (if
850 the queue_time_overall log selector is set), a one-byte patch to exigrep
851 was needed to allow it to recognize "Completed" as not the last thing in
854 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
855 patch that reportedly fixes this has been added. I am not expert enough to
856 create a test for it. This is what the patch creator wrote:
858 "I found a little strange behaviour of ldap code when working with
859 Windows 2003 AD Domain, where users was placed in more than one
860 Organization Units. When I tried to give exim partial DN, the exit code
861 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
862 But simultaneously result of request was absolutely normal ldap result,
863 so I produce this patch..."
865 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
866 REFERENCE, so I have modified the code to exclude the patch when that macro
869 55. Some experimental protocols are using DNS PTR records for new purposes. The
870 keys for these records are domain names, not reversed IP addresses. The
871 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
872 leaves it alone. Component reversal etc. now happens only for IP addresses.
875 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
877 57. Double the size of the debug message buffer (to 2048) so that more of very
878 long debug lines gets shown.
880 58. The exicyclog utility now does better if the number of log files to keep
881 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
883 59. Two changes related to the smtp_active_hostname option:
885 (1) $smtp_active_hostname is now available as a variable.
886 (2) The default for smtp_banner uses $smtp_active_hostname instead
887 of $primary_hostname.
889 60. The host_aton() function is supposed to be passed a string that is known
890 to be a valid IP address. However, in the case of IPv6 addresses, it was
891 not checking this. This is a hostage to fortune. Exim now panics and dies
892 if the condition is not met. A case was found where this could be provoked
893 from a dnsdb PTR lookup with an IPv6 address that had more than 8
894 components; fortuitously, this particular loophole had already been fixed
895 by change 4.50/55 above.
897 If there are any other similar loopholes, the new check in host_aton()
898 itself should stop them being exploited. The report I received stated that
899 data on the command line could provoke the exploit when Exim was running as
900 exim, but did not say which command line option was involved. All I could
901 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
905 61. There was a buffer overflow vulnerability in the SPA authentication code
906 (which came originally from the Samba project). I have added a test to the
907 spa_base64_to_bits() function which I hope fixes it.
910 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
911 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
913 63. The daemon start-up calls getloadavg() while still root for those OS that
914 need the first call to be done as root, but it missed one case: when
915 deliver_queue_load_max is set with deliver_drop_privilege. This is
916 necessary for the benefit of the queue runner, because there is no re-exec
917 when deliver_drop_privilege is set.
919 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
922 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
923 was in use, was not putting the data itself into the right store pool;
924 consequently, it could be overwritten for a subsequent message in the same
925 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
928 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
930 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
931 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
932 the function were treating the return as a boolean value, which happened to
933 work because 0=false and not-0=true, but is not correct code.
935 68. The host_aton() function was not handling scoped IPv6 addresses (those
936 with, for example, "%eth0" on the end) correctly.
938 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
939 negated items (that is, ~something) in unsigned ints. Some compilers
940 apparently mutter when there is no cast.
942 70. If an address verification called from an ACL failed, and did not produce a
943 user-specific message (i.e. there was only a "system" message), nothing was
944 put in $acl_verify_message. In this situation, it now puts the system
947 71. Change 4.23/11 added synchronization checking at the start of an SMTP
948 session; change 4.31/43 added the unwanted input to the log line - except
949 that it did not do this in the start of session case. It now does.
951 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
952 This is wrong and can cause the other end to generate a synchronization
953 error if it is another Exim or anything else that does the synchronization
954 check. A QUIT command is no longer sent after a timeout.
956 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
959 74. The defer_ok option of callout verification was not working if it was used
960 when verifying addresses in header lines, that is, for this case:
962 verify = header_sender/callout=defer_ok
964 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
965 those file descriptors could be used for SMTP connections. If anything
966 wrote to stderr (the example that came up was "warn" in embedded Perl), it
967 could be sent to the SMTP client, causing chaos. The daemon now opens
968 stdin, stdout, and stderr to /dev/null when it puts itself into the
971 76. Arrange for output from Perl's "warn" command to be written to Exim's main
972 log by default. The user can override this with suitable Perl magic.
974 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
975 the log message when discard triggers, was not working for the DATA ACL or
976 for the non-SMTP ACL.
978 78. Error message wording change in sieve.c.
980 79. If smtp_accept_max_per_host was set, the number of connections could be
981 restricted to fewer than expected, because the daemon was trying to set up
982 a new connection before checking whether the processes handling previous
983 connections had finished. The check for completed processes is now done
984 earlier. On busy systems, this bug wouldn't be noticed because something
985 else would have woken the daemon, and it would have reaped the completed
988 80. If a message was submitted locally by a user whose login name contained one
989 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
990 It caused a spool format error. I have fixed the spool reading code. A
991 related problem was that the "from" clause in the Received: line became
992 illegal because of the space(s). It is now covered by ${quote_local_part.
994 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
997 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
999 83. Patch from Sieve maintainer to fix unterminated string problem in
1000 "vacation" handling.
1002 84. Some minor changes to the Linux configuration files to help with other
1003 OS variants using glibc.
1005 85. One more patch for Sieve to update vacation handling to latest spec.
1008 ----------------------------------------------------
1009 See the note above about the 4.44 and 4.50 releases.
1010 ----------------------------------------------------
1016 1. Change 4.43/35 introduced a bug that caused file counts to be
1017 incorrectly computed when quota_filecount was set in an appendfile
1020 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1021 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1024 3. Give more explanation in the error message when the command for a transport
1025 filter fails to execute.
1027 4. There are several places where Exim runs a non-Exim command in a
1028 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1029 was being done only for the command run by the queryprogram router. It is
1030 now done for all such subprocesses. The other cases are: ${run, transport
1031 filters, and the commands run by the lmtp and pipe transports.
1033 5. Some older OS have a limit of 256 on the maximum number of file
1034 descriptors. Exim was using setrlimit() to set 1000 as a large value
1035 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1036 systems. I've change it so that if it can't get 1000, it tries for 256.
1038 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1039 was an oversight, and furthermore, ever since the addition of extra
1040 controls (e.g. 4.43/32), the checks on when to allow different forms of
1041 "control" were broken. There should now be diagnostics for all cases when a
1042 control that does not make sense is encountered.
1044 7. $recipients is now available in the predata ACL (oversight).
1046 8. Tidy the search cache before the fork to do a delivery from a message
1047 received from the command line. Otherwise the child will trigger a lookup
1048 failure and thereby defer the delivery if it tries to use (for example) a
1049 cached ldap connection that the parent has called unbind on.
1051 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1052 of $address_data from the recipient verification was clobbered by the
1053 sender verification.
1055 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1056 was its contents. (It was OK if the option was not defined at all.)
1058 11. A "Completed" log line is now written for messages that are removed from
1059 the spool by the -Mrm option.
1061 12. $host_address is now set to the target address during the checking of
1062 ignore_target_hosts.
1064 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1065 being passed; this would have caused $sender_host_name to have been used if
1066 matching the list had actually called for a host name (not very likely,
1067 since this list is usually IP addresses). A host name is now passed as
1070 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1071 code that specifies a non-restarting handler (typically sigaction() in
1072 modern systems) in an attempt to fix a rare and obscure crash bug.
1074 15. Narrowed the window for a race in the daemon that could cause it to ignore
1075 SIGCHLD signals. This is not a major problem, because they are used only to
1076 wake it up if nothing else does.
1078 16. A malformed maildirsize file could cause Exim to calculate negative values
1079 for the mailbox size or file count. Odd effects could occur as a result.
1080 The maildirsize information is now recalculated if the size or filecount
1083 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1084 support for a long time. Removed HAVE_SYS_VFS_H.
1086 18. Updated exipick to current release from John Jetmore.
1088 19. Allow an empty sender to be matched against a lookup in an address list.
1089 Previously the only cases considered were a regular expression, or an
1092 20. Exim went into a mad DNS lookup loop when doing a callout where the
1093 host was specified on the transport, if the DNS lookup yielded more than
1096 21. The RFC2047 encoding function was originally intended for short strings
1097 such as real names; it was not keeping to the 75-character limit for
1098 encoded words that the RFC imposes. It now respects the limit, and
1099 generates multiple encoded words if necessary. To be on the safe side, I
1100 have increased the buffer size for the ${rfc2047: expansion operator from
1103 22. Failure to deliver a bounce message always caused it to be frozen, even if
1104 there was an errors_to setting on the router. The errors_to setting is now
1107 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1108 canonical form (fully expanded) before being placed in
1109 $sender_host_address.
1111 24. Updated eximstats to version 1.33
1113 25. Include certificate and key file names in error message when GnuTLS fails
1114 to set them up, because the GnuTLS error message doesn't include the name
1115 of the failing file when there is a problem reading it.
1117 26. Expand error message when OpenSSL has problems setting up cert/key files.
1120 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1121 (this can affect the format of dates).
1123 28. exim_tidydb, when checking for the continued existence of a message for
1124 which it has found a message-specific retry record, was not finding
1125 messages that were in split spool directories. Consequently, it was
1126 deleting retry records that should have stayed in existence.
1128 29. eximstats updated to version 1.35
1129 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1130 1.35 - bugfix such that pie charts by volume are generated correctly
1132 30. The SPA authentication driver was not abandoning authentication and moving
1133 on to the next authenticator when an expansion was forced to fail,
1134 contradicting the general specification for all authenticators. Instead it
1135 was generating a temporary error. It now behaves as specified.
1137 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1138 (the order specifies the preference for clients). The order is now AES256,
1139 AES128, 3DES, ARCFOUR128.
1141 31. Small patch to Sieve code - explicitly set From: when generating an
1144 32. Exim crashed if a remote delivery caused a very long error message to be
1145 recorded - for instance if somebody sent an entire SpamAssassin report back
1146 as a large number of 550 error lines. This bug was coincidentally fixed by
1147 increasing the size of one of Exim's internal buffers (big_buffer) that
1148 happened as part of the Exiscan merge. However, to be on the safe side, I
1149 have made the code more robust (and fixed the comments that describe what
1152 33. Some experimental protocols are using DNS PTR records for new purposes. The
1153 keys for these records are domain names, not reversed IP addresses. The
1154 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1155 leaves it alone. Component reversal etc. now happens only for IP addresses.
1158 34. The host_aton() function is supposed to be passed a string that is known
1159 to be a valid IP address. However, in the case of IPv6 addresses, it was
1160 not checking this. This is a hostage to fortune. Exim now panics and dies
1161 if the condition is not met. A case was found where this could be provoked
1162 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1163 components; fortuitously, this particular loophole had already been fixed
1164 by change 4.50/55 or 4.44/33 above.
1166 If there are any other similar loopholes, the new check in host_aton()
1167 itself should stop them being exploited. The report I received stated that
1168 data on the command line could provoke the exploit when Exim was running as
1169 exim, but did not say which command line option was involved. All I could
1170 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1171 running as the user.
1174 35. There was a buffer overflow vulnerability in the SPA authentication code
1175 (which came originally from the Samba project). I have added a test to the
1176 spa_base64_to_bits() function which I hope fixes it.
1179 36. The daemon start-up calls getloadavg() while still root for those OS that
1180 need the first call to be done as root, but it missed one case: when
1181 deliver_queue_load_max is set with deliver_drop_privilege. This is
1182 necessary for the benefit of the queue runner, because there is no re-exec
1183 when deliver_drop_privilege is set.
1185 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1186 was in use, was not putting the data itself into the right store pool;
1187 consequently, it could be overwritten for a subsequent message in the same
1188 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1191 38. Sometimes the final signoff response after QUIT could fail to get
1192 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1193 < 0 before doing a fflush(). This bug looks as though it goes back to the
1194 introduction of TLS in release 3.20, but "sometimes" must have been rare
1195 because the tests only now provoked it.
1201 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1202 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1203 Now it returns an integer. A number of calls were still expecting a T/F
1204 return. Fortuitously, in all cases, the tests worked in OK situations,
1205 which is the norm. However, things would have gone wrong on any write
1206 failures on the smtp file descriptor. This function is used when sending
1207 messages over SMTP and also when doing verify callouts.
1209 2. When Exim is called to do synchronous delivery of a locally submitted
1210 message (the -odf or -odi options), it no longer closes stderr before doing
1213 3. Implemented the mua_wrapper option.
1215 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1217 5. Implemented the functions header_remove(), header_testname(),
1218 header_add_at_position(), and receive_remove_recipient(), and exported them
1221 6. If an ACL "warn" statement specified the addition of headers, Exim already
1222 inserted X-ACL-Warn: at the start if there was no header name. However, it
1223 was not making this test for the second and subsequent header lines if
1224 there were newlines in the string. This meant that an invalid header could
1225 be inserted if Exim was badly configured.
1227 7. Allow an ACL "warn" statement to add header lines at the start or after all
1228 the Received: headers, as well as at the end.
1230 8. Added the rcpt_4xx retry error code.
1232 9. Added postmaster_mailfrom=xxx to callout verification option.
1234 10. Added mailfrom=xxxx to the callout verification option, for verify=
1237 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1238 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1240 12. Inserted some casts to stop certain compilers warning when using pointer
1241 differences as field lengths or precisions in printf-type calls (mostly
1242 affecting debugging statements).
1244 13. Added optional readline() support for -be (dynamically loaded).
1246 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1247 same clock tick as a message's arrival, so that its received time was the
1248 same as the "first fail" time on the retry record, and that message
1249 remained on the queue past the ultimate address timeout, every queue runner
1250 would try a delivery (because it was past the ultimate address timeout) but
1251 after another failure, the ultimate address timeout, which should have then
1252 bounced the address, did not kick in. This was a "< instead of <=" error;
1253 in most cases the first failure would have been in the next clock tick
1254 after the received time, and all would be well.
1256 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1257 being recognized when the domain list was tested by the match_domain
1258 condition in an expansion string.
1260 16. Added the ${str2b64: operator.
1262 17. Exim was always calling setrlimit() to set a large limit for the number of
1263 processes, without checking whether the existing limit was already
1264 adequate. (It did check for the limit on file descriptors.) Furthermore,
1265 errors from getrlimit() and setrlimit() were being ignored. Now they are
1266 logged to the main and panic logs, but Exim does carry on, to try to do its
1267 job under whatever limits there are.
1269 18. Imported PCRE 5.0.
1271 19. Trivial typo in log message " temporarily refused connection" (the leading
1274 20. If the log selector return_path_on_delivery was set and an address was
1275 redirected to /dev/null, the delivery process crashed because it assumed
1276 that a return path would always be set for a "successful" delivery. In this
1277 case, the whole delivery is bypassed as an optimization, and therefore no
1280 21. Internal re-arrangement: the function for sending a challenge and reading
1281 a response while authentication was assuming a zero-terminated challenge
1282 string. It's now changed to take a pointer and a length, to allow for
1283 binary data in such strings.
1285 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1287 23. Exim was not respecting finduser_retries when seeking the login of the
1288 uid under which it was called; it was always trying 10 times. (The default
1289 setting of finduser_retries is zero.) Also, it was sleeping after the final
1290 failure, which is pointless.
1292 24. Implemented tls_on_connect_ports.
1294 25. Implemented acl_smtp_predata.
1296 26. If the domain in control=submission is set empty, Exim assumes that the
1297 authenticated id is a complete email address when it generates From: or
1298 Sender: header lines.
1300 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1301 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1302 chown and chgrp in /bin and hostname in /usr/bin.
1304 28. Exim was keeping the "process log" file open after each use, just as it
1305 does for the main log. This opens the possibility of it remaining open for
1306 long periods when the USR1 signal hits a daemon. Occasional processlog
1307 errors were reported, that could have been caused by this. Anyway, it seems
1308 much more sensible not to leave this file open at all, so that is what now
1311 29. The long-running daemon process does not normally write to the log once it
1312 has entered its main loop, and it closes the log before doing so. This is
1313 so that log files can straightforwardly be renamed and moved. However,
1314 there are a couple of unusual error situations where the daemon does write
1315 log entries, and I had neglected to close the log afterwards.
1317 30. The text of an SMTP error response that was received during a remote
1318 delivery was being truncated at 512 bytes. This is too short for some of
1319 the long messages that one sometimes sees. I've increased the limit to
1322 31. It is now possible to make retry rules that apply only when a message has a
1323 specific sender, in particular, an empty sender.
1325 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1326 it possible to be selective about when SMTP synchronization is enforced.
1328 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1330 32. Implemented hosts_connection_nolog.
1332 33. Added an ACL for QUIT.
1334 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1337 35. Added mailbox_size and mailbox_filecount to appendfile.
1339 36. Added control = no_multiline_responses to ACLs.
1341 37. There was a bug in the logic of the code that waits for the clock to tick
1342 in the case where the clock went backwards by a substantial amount such
1343 that the microsecond fraction of "now" was more than the microsecond
1344 fraction of "then" (but the whole seconds number was less).
1346 38. Added support for the libradius Radius client library this is found on
1347 FreeBSD (previously only the radiusclient library was supported).
1353 1. When certain lookups returned multiple values in the form name=value, the
1354 quoting of the values was not always being done properly. Specifically:
1355 (a) If the value started with a double quote, but contained no whitespace,
1357 (b) If the value contained whitespace other than a space character (i.e.
1358 tabs or newlines or carriage returns) it was not quoted.
1359 This fix has been applied to the mysql and pgsql lookups by writing a
1360 separate quoting function and calling it from the lookup code. The fix
1361 should probably also be applied to nisplus, ibase and oracle lookups, but
1362 since I cannot test any of those, I have not disturbed their existing code.
1364 2. A hit in the callout cache for a specific address caused a log line with no
1365 reason for rejecting RCPT. Now it says "Previous (cached) callout
1366 verification failure".
1368 3. There was an off-by-one bug in the queryprogram router. An over-long
1369 return line was truncated at 256 instead of 255 characters, thereby
1370 overflowing its buffer with the terminating zero. As well as fixing this, I
1371 have increased the buffer size to 1024 (and made a note to document this).
1373 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1374 when Exim is waiting for an SMTP response from a remote server, Exim
1375 restarts its select() call on the socket, thereby resetting its timeout.
1376 This is not a problem when such interrupts are rare. Somebody set up a cron
1377 job to run exiwhat every 2 minutes, which is less than the normal select()
1378 timeout (5 or 10 minutes). This meant that the select() timeout never
1379 kicked in because it was always reset. I have fixed this by comparing the
1380 time when an interrupt arrives with the time at the start of the first call
1381 to select(). If more time than the timeout has elapsed, the interrupt is
1382 treated as a timeout.
1384 5. Some internal re-factoring in preparation for the addition of Sieve
1385 extensions (by MH). In particular, the "personal" test is moved to a
1386 separate function, and given an option for scanning Cc: and Bcc: (which is
1387 not set for Exim filters).
1389 6. When Exim created an email address using the login of the caller as the
1390 local part (e.g. when creating a From: or Sender: header line), it was not
1391 quoting the local part when it contained special characters such as @.
1393 7. Installed new OpenBSD configuration files.
1395 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1396 try to make them clearer.
1398 9. Callout options, other than the timeout value, were being ignored when
1399 verifying sender addresses in header lines. For example, when using
1401 verify = header_sender/callout=no_cache
1403 the cache was (incorrectly) being used.
1405 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1406 only the Cygwin environment.
1408 11. When return_path_on_delivery was set as a log selector, if different remote
1409 addresses in the same message used different return paths and parallel
1410 remote delivery occurred, the wrong values would sometimes be logged.
1411 (Whenever a remote delivery process finished, the return path value from
1412 the most recently started remote delivery process was logged.)
1414 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1415 lines when AUTH and/or TLS are in use. This is the "received protocol"
1416 field. Exim used to use "asmtp" for authenticated SMTP, without any
1417 indication (in the protocol name) for TLS use. Now it follows the RFC and
1418 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1419 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1420 names appear in log lines as well as in Received: header lines.
1422 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1423 extensions, and comparison tests, and to fix some bugs.
1425 14. Changes to the "personal" filter test:
1427 (1) The test was buggy in that it was just doing the equivalent of
1428 "contains" tests on header lines. For example, if a user's address was
1429 anne@some.where, the "personal" test would incorrectly be true for
1431 To: susanne@some.where
1433 This test is now done by extracting each address from the header in turn,
1434 and checking the entire address. Other tests that are part of "personal"
1435 are now done using regular expressions (for example, to check local parts
1436 of addresses in From: header lines).
1438 (2) The list of non-personal local parts in From: addresses has been
1439 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1440 taken from the Sieve specification recommendations.
1442 (3) If the message contains any header line starting with "List-" it is
1443 treated as non-personal.
1445 (4) The test for "circular" in the Subject: header line has been removed
1446 because it now seems ill-conceived.
1448 15. Minor typos in src/EDITME comments corrected.
1450 16. Installed latest exipick from John Jetmore.
1452 17. If headers_add on a router specified a text string that was too long for
1453 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1454 of string_sprintf() is now avoided.
1456 18. $message_body_size was not set (it was always zero) when running the DATA
1457 ACL and the local_scan() function.
1459 19. For the "mail" command in an Exim filter, no default was being set for
1460 the once_repeat time, causing a random time value to be used if "once" was
1461 specified. (If the value happened to be <= 0, no repeat happened.) The
1462 default is now 0s, meaning "never repeat". The "vacation" command was OK
1463 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1464 (I found it when inspecting the code).
1466 20. There is now an overall timeout for performing a callout verification. It
1467 defaults to 4 times the callout timeout, which applies to individual SMTP
1468 commands during the callout. The overall timeout applies when there is more
1469 than one host that can be tried. The timeout is checked before trying the
1470 next host. This prevents very long delays if there are a large number of
1471 hosts and all are timing out (e.g. when the network connections are timing
1472 out). The value of the overall timeout can be changed by specifying an
1473 additional sub-option for "callout", called "maxwait". For example:
1475 verify = sender/callout=5s,maxwait=20s
1477 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1478 to the end before writing, but this should make it even safer).
1480 22. Exim was forgetting that it had advertised PIPELINING for the second and
1481 subsequent messages on an SMTP connection. It was also not resetting its
1482 memory on STARTTLS and an internal HELO.
1484 23. When Exim logs an SMTP synchronization error within a session, it now
1485 records whether PIPELINING has been advertised or not.
1487 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1488 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1489 rather than long int.
1491 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1493 26. Added the never_mail option to autoreply.
1499 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1500 crash if the getsockname() call failed; this can happen if a connection is
1501 closed very soon after it is established. The problem was simply in the
1502 order in which certain operations were done, causing Exim to try to write
1503 to the SMTP stream before it had set up the file descriptor. The bug has
1504 been fixed by making things happen in the correct order.
1510 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1511 before the connection was closed, thus losing the rejection response.
1513 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1514 some early Solaris releases, but causes trouble in current releases where
1515 socklen_t is defined.
1517 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1520 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1523 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1524 that releasing the top of it at the end releases what was used for sub-
1525 expansions (unless the block got too big). However, discard this block if
1526 the first thing is a variable or header, so that we can use its block when
1527 it is dynamic (useful for very large $message_headers, for example).
1529 6. Lookups now cache *every* query, not just the most recent. A new, separate
1530 store pool is used for this. It can be recovered when all lookup caches are
1531 flushed. Lookups now release memory at the end of their result strings.
1532 This has involved some general refactoring of the lookup sources.
1534 7. Some code has been added to the store_xxx() functions to reduce the amount
1535 of flapping under certain conditions.
1537 8. log_incoming_interface used to affect only the <= reception log lines. Now
1538 it causes the local interface and port to be added to several more SMTP log
1539 lines, for example "SMTP connection from", and rejection lines.
1541 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1543 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1545 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1546 could be overwritten at the end of the current message (or the start of a
1547 new message if it was set in a HELO ACL). The value is now preserved for
1548 the duration of the SMTP connection.
1550 12. If a transport had a headers_rewrite setting, and a matching header line
1551 contained an unqualified address, that address was qualified, even if it
1552 did not match any rewriting rules. The underlying bug was that the values
1553 of the flags that permit the existence of unqualified sender and recipient
1554 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1555 non-local messages, and by -bnq for local messages) were not being
1556 preserved with the message after it was received.
1558 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1559 "next input=" as part of the text comprising the host identity instead of
1560 the correct text. The code was using the same buffer for two different
1561 strings. However, depending on which order the printing function evaluated
1562 its arguments, the bug did not always show up. Under Linux, for example, my
1563 test suite worked just fine.
1565 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1566 doesn't work with some older versions of Perl. It has been changed to "my",
1567 which in any case is probably the better facility to use.
1569 15. A really picky compiler found some instances of statements for creating
1570 error messages that either had too many or two few arguments for the format
1573 16. The size of the buffer for calls to the DNS resolver has been increased
1574 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1575 for addresses that have a lot of PTR records. This alleviates a problem; it
1576 does not fully solve it.
1578 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1579 buffer now truncates the list and logs the incident, which is the same
1580 action as happens when Exim is looking up a host name and its aliases.
1581 Previously in this situation something unpredictable would happen;
1582 sometimes it was "internal error: store_reset failed".
1584 18. If a server dropped the connection unexpectedly when an Exim client was
1585 using GnuTLS and trying to read a response, the client delivery process
1586 crashed while trying to generate an error log message.
1588 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1589 string, for example:
1591 warn message = H1: something\nH2: something
1593 the text was added as a single header line from Exim's point of view
1594 though it ended up OK in the delivered message. However, searching for the
1595 second and subsequent header lines using $h_h2: did not work. This has been
1596 fixed. Similarly, if a system filter added multiple headers in this way,
1597 the routers could not see them.
1599 20. Expanded the error message when iplsearch is called with an invalid key to
1600 suggest using net-iplsearch in a host list.
1602 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1603 ACLs are no longer actually imposed (and a message to that effect is
1606 22. If a "gecos" field in a passwd entry contained escaped characters, in
1607 particular, if it contained a \" sequence, Exim got it wrong when building
1608 a From: or a Sender: header from that name. A second bug also caused
1609 incorrect handling when an unquoted " was present following a character
1610 that needed quoting.
1612 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1613 was not being matched caselessly.
1615 24. Arranged for all hyphens in the exim.8 source to be escaped with
1618 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1619 properly. Recipient callout cache records were still being keyed to include
1620 the sender, even when use_sender was set false. This led to far more
1621 callouts that were necessary. The sender is no longer included in the key
1622 when use_sender is false.
1624 26. Added "control = submission" modifier to ACLs.
1626 27. Added the ${base62d: operator to decode base 62 numbers.
1628 28. dnsdb lookups can now access SRV records.
1630 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1631 the configuration file.
1633 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1634 (-v) mode. This makes the output for a verbose queue run more intelligible.
1636 31. Added a use_postmaster feature to recipient callouts.
1638 32. Added the $body_zerocount variable, containing the number of binary zero
1639 bytes in the message body.
1641 33. The time of last modification of the "new" subdirectory is now used as the
1642 "mailbox time last read" when there is a quota error for a maildir
1645 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1647 35. Added +ignore_unknown as a special item in host lists.
1649 36. Code for decoding IPv6 addresses in host lists is now included, even if
1650 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1651 address was recognized as an IP address, but was then not correctly decoded
1652 into binary, causing unexpected and incorrect effects when compared with
1659 1. Very minor rewording of debugging text in manualroute to say "list of
1660 hosts" instead of "hostlist".
1662 2. If verify=header_syntax was set, and a header line with an unqualified
1663 address (no domain) and a large number of spaces between the end of the
1664 name and the colon was received, the reception process suffered a buffer
1665 overflow, and (when I tested it) crashed. This was caused by some obsolete
1666 code that should have been removed. The fix is to remove it!
1668 3. When running in the test harness, delay a bit after writing a bounce
1669 message to get a bit more predictability in the log output.
1671 4. Added a call to search_tidyup() just before forking a reception process. In
1672 theory, someone could use a lookup in the expansion of smtp_accept_max_
1673 per_host which, without the tidyup, could leave open a database connection.
1675 5. Added the variables $recipient_data and $sender_data which get set from a
1676 lookup success in an ACL "recipients" or "senders" condition, or a router
1677 "senders" option, similar to $domain_data and $local_part_data.
1679 6. Moved the writing of debug_print from before to after the "senders" test
1682 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1683 problems for message scanning, either using a data ACL, or using
1684 local_scan() because the Received: header was not generated till after they
1685 were called (in order to set the time as the time of reception completion).
1686 I have revised the way this works. The header is now generated after the
1687 body is received, but before the ACL or local_scan() are called. After they
1688 are run, the timestamp in the header is updated.
1694 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1695 before starting a queue runner without re-exec. This happened only when
1696 deliver_drop_privilege was set or when the Exim user was set to root. The
1697 effect of the bug was that timeouts during subsequent deliveries caused
1698 crashes instead of being properly handled. The handler is now left at its
1699 default (and expected) setting.
1701 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1702 message, again when deliver_drop_privilege is set or Exim is run as root.
1703 The bug described in (1) was not present in this case, but the tidying up
1704 of the other signals was missing. I have made the two cases consistent.
1706 3. The ignore_target_hosts setting on a manualroute router was being ignored
1707 for hosts that were looked up using the /MX notation.
1709 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1712 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1713 operated on the sender address. After changing the $sender_address to <>
1714 for the sender address verify, Exim was re-instated it as the original
1715 (before rewriting) address, but remembering that it had rewritten it, so it
1716 wasn't rewriting it again. This bug also had the effect of breaking the
1717 sender address verification caching when the sender address was rewritten.
1719 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1720 This has been changed so that if the ip literal address matches
1721 ignore_target_hosts, the router declines.
1723 7. Added expansion conditions match_domain, match_address, and match_local_
1724 part (NOT match_host).
1726 8. The placeholder for the Received: header didn't have a length field set.
1728 9. Added code to Exim itself and to exim_lock to test for a specific race
1729 condition that could lead to file corruption when using MBX delivery. The
1730 issue is with the lockfile that is created in /tmp. If this file is removed
1731 after a process has opened it but before that process has acquired a lock,
1732 there is the potential for a second process to recreate the file and also
1733 acquire a lock. This could lead to two Exim processes writing to the file
1734 at the same time. The added code performs the same test as UW imapd; it
1735 checks after acquiring the lock that its file descriptor still refers to
1736 the same named file.
1738 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1739 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1742 11. Added the smtp_active_hostname option. If used, this will typically be made
1743 to depend on the incoming interface address. Because $interface_address is
1744 not set up until the daemon has forked a reception process, error responses
1745 that can happen earlier (such as "too many connections") no longer contain
1748 12. If an expansion in a condition on a "warn" statement fails because a lookup
1749 defers, the "warn" statement is abandoned, and the next ACL statement is
1750 processed. Previously this caused the whole ACL to be aborted.
1752 13. Added the iplsearch lookup type.
1754 14. Added ident_timeout as a log selector.
1756 15. Added tls_certificate_verified as a log selector.
1758 16. Added a global option tls_require_ciphers (compare the smtp transport
1759 option of the same name). This controls incoming TLS connections.
1761 17. I finally figured out how to make tls_require_ciphers do a similar thing
1762 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1763 before starting the TLS session.
1765 18. Tabs are now shown as \t in -bP output.
1767 19. If the log selector return_path_on_delivery was set, Exim crashed when
1768 bouncing a message because it had too many Received: header lines.
1770 20. If two routers both had headers_remove settings, and the first one included
1771 a superfluous trailing colon, the final name in the first list and the
1772 first name in the second list were incorrectly joined into one item (with a
1773 colon in the middle).
1779 1. Added -C and -D options to the exinext utility, mainly to make it easier
1780 to include in the automated testing, but these could be helpful when
1781 multiple configurations are in use.
1783 2. The exinext utility was not formatting the output nicely when there was
1784 an alternate port involved in the retry record key, nor when there was a
1785 message id as well (for retries that were specific to a specific message
1786 and a specific host). It was also confused by IPv6 addresses, because of
1787 the additional colons they contain. I have fixed the IPv4 problem, and
1788 patched it up to do a reasonable job for IPv6.
1790 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1791 delivery, the log line now contains "pipelined" if PIPELINING was used.
1793 4. An SMTP transport process used to panic and die if the bind() call to set
1794 an explicit outgoing interface failed. This has been changed; it is now
1795 treated in the same way as a connect() failure.
1797 5. A reference to $sender_host_name in the part of a conditional expansion
1798 that was being skipped was still causing a DNS lookup. This no longer
1801 6. The def: expansion condition was not recognizing references to header lines
1802 that used bh_ and bheader_.
1804 7. Added the _cache feature to named lists.
1806 8. The code for checking quota_filecount in the appendfile transport was
1807 allowing one more file than it should have been.
1809 9. For compatibility with Sendmail, the command line option
1817 and sets the incoming protocol and host name (for trusted callers). The
1818 host name and its colon can be omitted when only the protocol is to be set.
1819 Note the Exim already has two private options, -pd and -ps, that refer to
1820 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1821 "s", but I don't think that's a major issue.
1823 10. A number of refactoring changes to the code, none of which should affect
1826 (a) The number of logging options was getting close to filling up the
1827 32-bit word that was used as a bit map. I have split them into two classes:
1828 those that are passed in the argument to log_write(), and those that are
1829 only ever tested independently outside of that function. These are now in
1830 separate 32-bit words, so there is plenty of room for expansion again.
1831 There is no change in the user interface or the logging behaviour.
1833 (b) When building, for example, log lines, the code previously used a
1834 macro that called string_cat() twice, in order to add two strings. This is
1835 not really sufficiently general. Furthermore, there was one instance where
1836 it was actually wrong because one of the argument was used twice, and in
1837 one call a function was used. (As it happened, calling the function twice
1838 did not affect the overall behaviour.) The macro has been replaced by a
1839 function that can join an arbitrary number of extra strings onto a growing
1842 (c) The code for expansion conditions now uses a table and a binary chop
1843 instead of a serial search (which was left over from when there were very
1844 few conditions). Also, it now recognizes conditions like "pam" even when
1845 the relevant support is not compiled in: a suitably worded error message is
1846 given if an attempt is made to use such a condition.
1848 11. Added ${time_interval:xxxxx}.
1850 12. A bug was causing one of the ddress fields not to be passed back correctly
1851 from remote delivery subprocesses. The field in question was not being
1852 subsequently used, so this caused to problems in practice.
1854 13. Added new log selectors queue_time and deliver_time.
1856 14. Might have fixed a bug in maildirsizefile handling that threw up
1857 "unexpected character" debug warnings, and recalculated the data
1858 unnecessarily. In any case, I expanded the warning message to give more
1861 15. Added the message "Restricted characters in address" to the statements in
1862 the default ACL that block characters like @ and % in local parts.
1864 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1865 Three changes have been made:
1867 (a) There was a serious bug; a negative response to MAIL caused the whole
1868 recipient domain to be cached as invalid, thereby blocking all messages
1869 to all local parts at the same domain, from all senders. This bug has
1870 been fixed. The domain is no longer cached after a negative response to
1871 MAIL if the sender used is not empty.
1873 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1876 (c) A new callout option, "use_sender" has been added for people who want
1877 the modified behaviour.
1883 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1886 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1887 indeed breaks things for older releases.
1889 3. Added additional logging to the case where there is a problem reading data
1890 from a filter that is running in a subprocess using a pipe, in order to
1891 try to track down a specific problem.
1893 4. Testing facility fudge: when running in the test harness and attempting
1894 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1895 getting "No route to host". Convert this to a timeout.
1897 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1900 6. Some OS don't have socklen_t but use size_t instead. This affects the
1901 fifth argument of getsockopt() amongst other things. This is now
1902 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1903 can be set for individual OS. I have set it for SunOS5, OSF1, and
1904 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1905 some earlier ones do not.
1907 7. Change 4.30/15 was not doing the test caselessly.
1909 8. The standard form for an IPv6 address literal was being rejected by address
1910 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1911 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1912 this, as well as the form without the "IPv6" on the front (but only when
1913 address literals are enabled, of course).
1915 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1917 10. Exim crashed if a message with an empty sender address specified by -f
1918 encountered a router with an errors_to setting. This could be provoked only
1919 by a command such as
1923 where an empty string was supplied; "<>" did not hit this bug.
1925 11. Installed PCRE release 4.5.
1927 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1928 remained set. It is now erased.
1930 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1931 times from message ids (which are base 36 rather than the normal 62).
1933 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1934 were being counted as actual protocol errors, and logged if the log
1935 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1936 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1937 and DATA following a set of rejected RCPTs do not count as protocol errors.
1938 In other words, Exim assumes they were pipelined, though this may not
1939 actually be the case. Of course, in all cases the client gets an
1940 appropriate error code.
1942 15. If a lookup fails in an ACL condition, a message about the failure may
1943 be available; it is used if testing the ACL cannot continue, because most
1944 such messages specify what the cause of the deferral is. However, some
1945 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1946 that caused an old message to be retained and used if a later statement
1947 caused a defer, replacing the real cause of the deferral.
1949 16. If an IP address had so many PTR records that the DNS lookup buffer
1950 was not large enough to hold them, Exim could crash while trying to process
1951 the truncated data. It now detects and logs this case.
1953 17. Further to 4.21/58, another change has been made: if (and only if) the
1954 first line of a message (the first header line) ends with CRLF, a bare LF
1955 in a subsequent header line has a space inserted after it, so as not to
1956 terminate the header.
1958 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1959 unnecessarily, used atoi() instead of strtol(), and didn't check the
1960 termination when getting file sizes from file names by regex.
1962 19. Completely re-implemented the support for maildirsize files, in the light
1963 of a number of problems with the previous contributed implementation
1964 (4.30/29). In particular:
1966 . If the quota is zero, the maildirsize file is maintained, but no quota is
1969 . If the maildir directory does not exist, it is created before any attempt
1970 to write a maildirsize file.
1972 . The quota value in the file is just a cache; if the quota is changed in
1973 the transport, the new value overrides.
1975 . A regular expression is available for excluding directories from the
1978 20. The autoreply transport checks the characters in options that define the
1979 message's headers; it allows continued headers, but it was checking with
1980 isspace() after an embedded newline instead of explicitly looking for a
1983 21. If all the "regular" hosts to which an address was routed had passed their
1984 expiry times, and had not reached their retry times, the address was
1985 bounced, even if fallback hosts were defined. Now Exim should go on to try
1988 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1989 equivalent code in the SMTP transport. Some hosts send humungous responses
1990 to HELO/EHLO, more than 1024 it seems.
1992 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1993 gives compiler warnings in some environments. I've now done it "properly",
1996 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1997 (because of problems with the built-in one) was declared to return uschar *
1998 instead of char *, causing compiler failure.
2000 25. Fixed a file descriptor leak when processing alias/forward files.
2002 26. Fixed a minor format string issue in dbfn.c.
2004 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2006 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2007 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2010 29. When a sender address is verified, it is cached, to save repeating the test
2011 when there is more than one recipient in a message. However, when the
2012 verification involves a callout, it is possible for different callout
2013 options to be set for different recipients. It is too complicated to keep
2014 track of this in the cache, so now Exim always runs a verification when a
2015 callout is required, relying on the callout cache for the optimization.
2016 The overhead is duplication of the address routing, but this should not be
2019 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2020 to be verified with callout=postmaster, and the main callout worked but the
2021 postmaster check failed, the verification correctly failed. However, if a
2022 subsequent RCPT command asked for sender verification *without* the
2023 postmaster check, incorrect caching caused this verification also to fail,
2026 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2027 it was not caching the DNS options (qualify_single, search_parents) that
2028 were used when the lookup failed. A subsequent lookup with different
2029 options therefore always gave the same answer, though there were cases
2030 where it should not have. (Example: a "domains = !$mx_any" option on a
2031 dnslookup router: the "domains" option is always processed without any
2032 widening, but the router might have qualify_single set.) Now Exim uses the
2033 cached value only when the same options are set.
2035 32. Added John Jetmore's "exipick" utility to the distribution.
2037 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2038 than a timeout (e.g. a certificate is required, and is not provided), an
2039 Exim server now closes the connection immediately. Previously it waited for
2040 the client to close - but if the client is SSL, it seems that they each
2041 wait for each other, leading to a delay before one of them times out.
2043 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2044 maintained 0.8.x compatibility because I don't think many are using it, and
2045 it is clearly obsolete.
2047 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2050 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2051 client certificate was expired. A simple patch fixes this, though I don't
2052 understand the full logic of why the verify callback is called multiple
2055 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2056 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2057 which causes problems with some clients (such as the Certicom SSL Plus
2058 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2059 disables the coutermeasure allowing Eudora to connect."
2061 38. Exim was not checking that a write() to a log file succeeded. This could
2062 lead to Bad Things if a log got too big, in particular if it hit a file
2063 size limit. Exim now panics and dies if it cannot write to a log file, just
2064 as it does if it cannot open a log file.
2066 39. Modified OS/Makefile-Linux so that it now contains
2068 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2070 The two -D definitions ensure that Exim is compiled with large file
2071 support, which makes it possible to handle log files that are bigger than
2074 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2075 instance) a domain was checked against a named list that involved a lookup,
2076 causing $domain_data to be set, then another domain was checked against the
2077 same list, then the first domain was re-checked, the value of $domain_data
2078 after the final check could be wrong. In particular, if the second check
2079 failed, it could be set empty. This bug probably also applied to
2082 41. The strip_trailing_dot option was not being applied to the address given
2083 with the -f command-line option.
2085 42. The code for reading a message's header from the spool was incrementing
2086 $received_count, but never initializing it. This meant that the value was
2087 incorrect (doubled) while delivering a message in the same process in which
2088 it was received. In the most common configuration of Exim, this never
2089 happens - a fresh exec is done - but it can happen when
2090 deliver_drop_privilege is set.
2092 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2093 it now includes up to 150 characters of the unexpected data in the log
2096 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2097 and building data strings. The size of both of these buffers was 10 000
2098 bytes - far larger than anybody would *ever* want, thought I. Needless to
2099 say, somebody hit the limit. I have increased the maximum line length to
2100 20 000 and the maximum data length of concatenated lines to 100 000. I have
2101 also fixed two bugs, because there was no checking on these buffers. Tsk,
2102 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2103 buffer is too small.
2105 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2106 lsearch lookups. Now it does.
2108 46. When parsing a route_list item in a manualroute router, a fixed-length
2109 buffer was used for the list of hosts. I made this 1024 bytes long,
2110 thinking that nobody would ever have a list of hosts that long. Wrong.
2111 Somebody had a whole pile of complicated expansion conditions, and the
2112 string was silently truncated, leading to an expansion error. It turns out
2113 that it is easier to change to an unlimited length (owing to other changes
2114 that have happened since this code was originally written) than to build
2115 structure for giving a limitation error. The length of the item that
2116 expands into the list of hosts is now unlimited.
2118 47. The lsearch lookup could not handle data where the length of text line was
2119 more than 4095 characters. Such lines were truncated, leading to shortened
2120 data being returned. It should now handle lines of any length.
2122 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2123 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2126 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2127 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2128 no longer happens in comment lines. A list of replacements is now placed
2129 at the head of all of the source files, except those whose only change is
2130 to replace PERL_COMMAND in the very first #! line.
2132 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2133 messages on the queue, with a bottom-up merge sort, using code contributed
2134 by Michael Haardt. This should make operations like -bp somewhat faster on
2135 large queues. It won't affect queue runners, except when queue_run_in_order
2138 51. Installed eximstats 1.31 in the distribution.
2140 52. Added support for SRV lookups to the dnslookup router.
2142 53. If an ACL referred to $message_body or $message_body_end, the value was not
2143 reset for any messages that followed in the same SMTP session.
2145 54. The store-handling optimization for building very long strings was not
2146 differentiating between the different store pools. I don't think this
2147 actually made any difference in practice, but I've tidied it.
2149 55. While running the routers to verify a sender address, $sender_address
2150 was still set to the sender address. This is wrong, because when routing to
2151 send a bounce to the sender, it would be empty. Therefore, I have changed
2152 it so that, while verifying a sender address, $sender_address is set to <>.
2153 (There is no change to what happens when verifying a recipient address.)
2155 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2156 target A or AAAA records (if not already returned) without resetting the
2157 qualify_single or search_parents options of the DNS resolver. These are
2158 inappropriate in this case because the targets of MX and SRV records must
2159 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2160 target that, when qualified, matched something in the local domain. These
2161 two options are now turned off when doing these lookups.
2163 57. It seems that at least some releases of Reiserfs (which does not have the
2164 concept of a fixed number of inodes) returns zero and not -1 for the
2165 number of available inodes. This interacted badly with check_spool_inodes,
2166 which assumed that -1 was the "no such thing" setting. What I have done is
2167 to check that the total number of inodes is greater than zero before doing
2168 the test of how many are available.
2170 58. When a "warn" ACL statement has a log_message modifier, the message is
2171 remembered, and not repeated. This is to avoid a lot of repetition when a
2172 message has many recipients that cause the same warning to be written.
2173 Howewer, Exim was preserving the list of already written lines for an
2174 entire SMTP session, which doesn't seem right. The memory is now reset if a
2175 new message is started.
2177 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2178 debugging output unless log_rewrite was also set.
2180 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2181 of (int)(handle) when we know that handle contains (void *)(-1).
2183 61. The Exim daemon panic-logs an error return when it closes the incoming
2184 connection. However "connection reset by peer" seems to be common, and
2185 isn't really an error worthy of noting specially, so that particular error
2188 62. When Exim is trying to find all the local interfaces, it used to panic and
2189 die if the ioctl to get the interface flags failed. However, it seems that
2190 on at least one OS (Solaris 9) it is possible to have an interface that is
2191 included in the list of interfaces, but for which you get a failure error
2192 for this call. This happens when the interface is not "plumbed" into a
2193 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2194 failure of the "get flags" call assumes that the interface is down.
2196 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2197 makes life easier for people who are doing arithmetic on fields extracted
2198 from dates, where you often get leading zeros that should not be
2199 interpreted as octal.
2201 64. Added qualify_domain to the redirect router, to override the global
2204 65. If a pathologically long header line contained very many addresses (the
2205 report of this problem mentioned 10 000) and each of them was rewritten,
2206 Exim could use up a very large amount of memory. (It kept on making new
2207 copies of the header line as it rewrote, and never released the old ones.)
2208 At the expense of a bit more processing, the header rewriting function has
2209 been changed so that it no longer eats memory in this way.
2211 66. The generation of the Received: header has been moved from the time that a
2212 message starts to be received, to the time that it finishes. The timestamp
2213 in the Received: header should now be very close to that of the <= log
2214 line. There are two side-effects of this change:
2216 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2217 logged header lines no longer include the local Received: line, because
2218 it has not yet been created. The same applies to a copy of the message
2219 that is returned to a non-SMTP sender when a message is rejected.
2221 (b) When a filter file is tested using -bf, no additional Received: header
2222 is added to the test message. After some thought, I decided that this
2225 This change does not affect the value of $received_for. It is still set
2226 after address rewriting, but before local_scan() is called.
2228 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2230 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2231 gave an unhelpful panic error message, and a defer error. I have managed to
2232 change this behaviour so that it now rejects any supplied certificate,
2233 which seems right, as the list of acceptable certificates is empty.
2235 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2236 gave an unhelpful defer error. I have not managed to make this reject any
2237 supplied certificates, but the error message it gives is "no certificate
2238 supplied", which is not helpful.
2240 70. exigrep's output now also includes lines that are not associated with any
2241 message, but which match the given pattern. Implemented by a patch from
2242 Martin Sluka, which also tidied up the Perl a bit.
2244 71. Recipient callout verification, like sender verification, was using <> in
2245 the MAIL FROM command. This isn't really the right thing, since the actual
2246 sender may affect whether the remote host accepts the recipient or not. I
2247 have changed it to use the actual sender in the callout; this means that
2248 the cache record is now keyed on a recipient/sender pair, not just the
2249 recipient address. There doesn't seem to be a real danger of callout loops,
2250 since a callout by the remote host to check the sender would use <>.
2251 [SEE ABOVE: changed after hitting problems.]
2253 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2254 temporary errors. However, in the case of such a code being given after
2255 the end of a data transmission (i.e. after ".") Exim was failing to write
2256 a retry record for the message. (Yes, there was some broken host that was
2257 actually sending 8xx at this point.)
2259 73. An unknown lookup type in a host list could cause Exim to panic-die when
2260 the list was checked. (An example that provoked this was putting <; in the
2261 middle of a list instead of at the start.) If this happened during a DATA
2262 ACL check, a -D file could be left lying around. This kind of configuration
2263 error no longer causes Exim to die; instead it causes a defer errror. The
2264 incident is still logged to the main and panic logs.
2266 74. Buglet left over from Exim 3 conversion. The message "too many messages
2267 in one connection" was written to the rejectlog but not the mainlog, except
2268 when address rewriting (yes!) was being logged.
2270 75. Added write_rejectlog option.
2272 76. When a system filter was run not as root (that is, when system_filter_user
2273 was set), the values of the $n variables were not being returned to the
2274 main process; thus, they were not subsequently available in the $sn
2277 77. Added +return_path_on_delivery log selector.
2279 78. A connection timeout was being treated differently from recipients deferred
2280 when testing hosts_max_try with a message that was older than the host's
2281 retry timeout. (The host should not be counted, thus allowing all hosts to
2282 be tried at least once before bouncing.) This may have been the cause of an
2283 occasionally reported bug whereby a message would remain on the queue
2284 longer than the retry timeout, but would be bounced if a delivery was
2285 forced. I say "may" because I never totally pinned down the problem;
2286 setting up timeout/retry tests is difficult. See also the next item.
2288 79. The ultimate address timeout was not being applied to errors that involved
2289 a combination of host plus message (for example, a timeout on a MAIL
2290 command). When an address resolved to a number of possible hosts, and they
2291 were not all tried for each delivery (e.g. because of hosts_max_try), a
2292 message could remain on the queue longer than the retry timeout.
2294 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2297 81. Fixed an obscure SMTP outgoing bug which required at least the following
2298 conditions: (a) there was another message waiting for the same server;
2299 (b) the server returned 5xx to all RCPT commands in the first message so
2300 that the message was not completed; (c) the server dropped the connection
2301 or gave a negative response to the RSET that Exim sends to abort the
2302 transaction. The observed case was a dropped connection after DATA that had
2303 been sent in pipelining mode. That is, the server had advertised PIPELINING
2304 but was not implementing it correctly. The effect of the bug was incorrect
2305 behaviour, such as trying another host, and this could lead to a crash.
2311 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2312 and daemon.c were passed as pointers to ints; they should have been
2313 pointers to socklen_t variables (which are typically unsigned ints).
2315 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2318 3. Fixed a really odd bug that affected only the testing scheme; patching a
2319 certain fixed string in the binary changed the value of another string that
2320 happened to be identical to the end of the original first string.
2322 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2323 name", it returns that address as the IP address. On some operating
2324 systems (e.g. Solaris), it also passes back the IP address string as the
2325 "host name". However, on others (e.g. Linux), it passes back an empty
2326 string. Exim wasn't checking for this, and was changing the host name to an
2327 empty string, assuming it had been canonicized.
2329 5. Although rare, it is permitted to have more than one PTR record for a given
2330 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2331 all the names associated with an address, because they do in Solaris.
2332 However, it seems that they do not in Linux for data that comes from the
2333 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2334 I found this out when I moved to a new Linux workstation and tried to run
2335 the Exim test suite.
2337 To get round this problem I have changed the code so that it now does its
2338 own call to the DNS to look up PTR records when searching for a host name.
2339 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2340 addresses that are only in /etc/hosts are still found.
2342 This behaviour is, however, controlled by an option called host_lookup_
2343 order, which defaults to "bydns:byaddr". If people want to use the other
2344 order, or indeed, just use one or the other means of lookup, they can
2345 specify it in this variable.
2347 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2348 some operating systems, this comes back from gethostbyaddr() as an empty
2349 string, and this is what Exim used to test for. However, it seems that in
2350 other systems, "." is yielded. Exim now tests for this case too.
2352 7. The values of check_spool_space and check_log_space are now held internally
2353 as a number of kilobytes instead of an absolute number of bytes. If a
2354 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2355 kilobyte. This means that much larger values can be stored.
2357 8. Exim monitor: an attempt to get the action menu when not actually pointing
2358 at a message produces an empty menu entitled "No message selected". This
2359 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2360 no entries in it ("Shell widget menu has zero width and/or height"). So I
2361 have added a single, blank menu entry in this case.
2363 9. Added ${quote_local_part.
2365 10. MIME decoding is now applied to the contents of Subject: header lines when
2368 11. Now that a reference to $sender_host_address automatically causes a reverse
2369 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2370 host lookup before query-style lookups in lists that might use this
2371 variable. This has therefore been abolished, and the "net-" prefix is no
2372 longer necessary for query-style lookups.
2374 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2375 have been a typo for LFLAGS, so it has been changed.
2377 13. The install script calls Exim with "-C /dev/null" in order to find the
2378 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2379 to be output. Howeve, since Exim outputs its version number before the
2380 error, it didn't break the script. It just looked ugly. I fixed this by
2381 always allowing "-C /dev/null" if the caller is root.
2383 14. Ignore overlarge ACL variable number when reading spool file - insurance
2384 against a later release with more variables having written the file.
2386 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2387 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2388 this, as well as the form without the "IPv6" on the front.
2390 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2391 OS/Makefile-Darwin file.
2393 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2394 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2396 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2397 variable. However, some OS use ints for inodes. I've added cast to long int
2398 to get rid of the compiler warning.
2400 19. I had forgotten to lock out "/../" in configuration file names when
2401 ALT_CONFIG_PREFIX was set.
2403 20. Routers used for verification do not need to specify transports. However,
2404 if such a router generated a host list, and callout was configured, Exim
2405 crashed, because it could not find a port number from the (non-existent)
2406 transport. It now assumes port 25 in this circumstance.
2408 21. Added the -t option to exigrep.
2410 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2411 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2412 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2413 them is set, LOOKUP_LSEARCH is forced.
2415 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2417 24. Added sender and host information to the "rejected by local_scan()" log
2418 line; previously there was no indication of these.
2420 25. Added .include_if_exists.
2422 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2423 Linux. It turns out that not all file systems support this. Apparently some
2424 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2425 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2426 not supported on the file descriptor, is now ignored when Exim is trying to
2427 sync a directory. This applies only to Linux.
2429 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2431 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2432 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2435 29. Added support for maildirsize files from supplied patch (modified a bit).
2437 30. The use of :fail: followed by an empty string could lead Exim to respond to
2438 sender verification failures with (e.g.):
2440 550 Verification failed for <xxx>
2441 550 Sender verify failed
2443 where the first response line was missing the '-' that indicates it is not
2444 the final line of the response.
2446 31. The loop for finding the name of the user that called Exim had a hardwired
2447 limit of 10; it now uses the value of finduser_retries, which is used for
2448 all other user lookups.
2450 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2453 33. Exim was neglecting to zero errno before one call of strtol() when
2454 expanding a string and expecting an integer value. On some systems this
2455 resulted in spurious "integer overflow" errors. Also, it was casting the
2456 result into an int without checking.
2458 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2459 did not work. The code looks as if it has *never* worked, though it appears
2460 to have been documented since at least releast 1.62. I have made it work.
2462 35. The "timeout_DNS" error in retry rules, also documented since at least
2463 1.62, also never worked. As it isn't clear exactly what this means, and
2464 clearly it isn't a major issue, I have abolished the feature by treating it
2465 as "timeout", and writing a warning to the main and panic logs.
2467 36. The display of retry rules for -brt wasn't always showing the error code
2470 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2471 timeout_connect_A, timeout_connect_MX.
2473 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2474 to the empty sender.
2476 39. The daemon was not analysing the content of -oX till after it had closed
2477 stderr and disconnected from the controlling terminal. This meant that any
2478 syntax errors were only noted on the panic log, and the return code from
2479 the command was 0. By re-arranging the code a little, I've made the
2480 decoding happen first, so such errors now appear on stderr, and the return
2481 code is 1. However, the actual setting up of the sockets still happens in
2482 the disconnected process, so errors there are still only recorded on the
2485 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2486 connections (as happens on some IP stacks) was logged at start up time as
2487 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2488 it from "IPv6 and IPv4", which means that two separate sockets are being
2491 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2492 says whether AF_INET or AF_INET6 was passed as an argument.
2494 42. Exiwhat output was messed up when time zones were included in log
2497 43. Exiwhat now gives more information about the daemon's listening ports,
2498 and whether -tls-on-connect was used.
2500 44. The "port" option of the smtp transport is now expanded.
2502 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2503 silently ignored. Now an error message is written to the main and panic
2506 46. There's a new ACL modifier called "logwrite" which writes to a log file
2507 as soon as it is encountered.
2509 47. Added $local_user_uid and $local_user_gid at routing time.
2511 48. Exim crashed when trying to verify a sender address that was being
2514 49. Exim was recognizing only a space character after ".include". It now also
2515 recognizes a tab character.
2517 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2518 extracting the relevant information from the specification. The man page no
2519 longer contains scrambled data for the -d option, and I've added a section
2520 at the front about calling Exim under different names.
2522 51. Added "extra_headers" argument to the "mail" command in filter files.
2524 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2527 53. Installed eximstats 1.29.
2529 54. Added transport_filter_timeout as a generic transport option.
2531 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2532 Cc: header lines. This was required by RFC 822, but it not required by RFC
2535 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2536 incoming messages that did not have them. Now it does so only if the
2537 message originates locally, that is, if there is no associated remote host
2538 address. When Resent- header lines are present, this applies to the Resent-
2539 lines rather than the non-Resent- lines.
2541 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2542 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2544 58. Messages for configuration errors now include the name of the main
2545 configuration file - useful now that there may be more than one file in a
2546 list (.included file names were always shown).
2548 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2549 for those rare installations that do not start the daemon as root or run it
2550 setuid root. I've cut out the call to initgroups() if the daemon is not
2553 60. The Exim user and group can now be bound into the binary as text strings
2554 that are looked up at the start of Exim's processing.
2556 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2558 62. Added $mailstore_basename variable.
2560 63. Installed patch to sieve.c from Michael Haardt.
2562 64. When Exim failed to open the panic log after failing to open the main log,
2563 the original message it was trying to log was written to stderr and debug
2564 output, but if they were not available (the usual case in production), it
2565 was lost. Now it is written to syslog before the two lines that record the
2566 failures to open the logs.
2568 65. Users' Exim filters run in subprocesses under the user's uid. It is
2569 possible for a "deliver" command or an alias in a "personal" command to
2570 provoke an address rewrite. If logging of address rewriting is configured,
2571 this fails because the process is not running as root or exim. There may be
2572 a better way of dealing with this, but for the moment (because 4.30 needs
2573 to be released), I have disabled address rewrite logging when running a
2574 filter in a non-root, non-exim process.
2580 1. The buildconfig auxiliary program wasn't quoting the value set for
2581 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2582 not defined. This bug was masked in 4.22 by the effect that was fixed in
2585 2. Some messages that were rejected after a message id was allocated were
2586 shown as "incomplete" by exigrep. It no longer does this for messages that
2587 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2589 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2590 have allow_domain_literals set, the ID did not get logged in the <= line.
2591 Domain literals are now always recognized in Message-ID: header lines.
2593 4. The first argument for a ${extract expansion item is the key name or field
2594 number. Leading and trailing spaces in this item were not being ignored,
2595 causing some misleading effects.
2597 5. When deliver_drop_privilege was set, single queue runner processes started
2598 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2599 same command in the process it spins off) were not dropping privilege.
2601 6. When the daemon running as "exim" started a queue runner, it always
2602 re-executed Exim in the spun-off process. This is a waste of effort when
2603 deliver_drop_privilege is set. The new process now just calls the
2604 queue-runner function directly.
2610 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2613 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2614 ignored. Though the use of -oP was forcing the writing of a pid file, it
2615 was always written to the default place.
2617 3. If the message "no IP address found for host xxxx" is generated during
2618 incoming verification, it is now followed by identification of the incoming
2619 connection (so you can more easily find what provoked it).
2621 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2623 5. Added some features to "harden" Exim a bit more against certain attacks:
2625 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2626 be put in Local/Makefile. This is like the never_users runtime option,
2627 but it cannot be overridden. The default setting is "root".
2629 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2630 prefix string with which any file named in a -C command line option
2633 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2634 is retained for -C and -D only if the caller of Exim is root. Without
2635 it, the exim user may also use -C and -D and retain privilege.
2637 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2638 command line option is disabled.
2640 6. Macro names set by the -D option must start with an upper case letter, just
2641 like macro names defined in the configuration file.
2643 7. Added "dereference=" facility to LDAP.
2645 8. Two instances of the typo "uknown" in the source files are fixed.
2647 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2648 the Configure-Makefile script screwed up while processing it.
2650 10. Incorporated PCRE 4.4.
2652 11. The SMTP synchronization check was not operating right at the start of an
2653 SMTP session. For example, it could not catch a HELO sent before the client
2654 waited for the greeting. There is now a check for outstanding input at the
2655 point when the greeting is written. Because of the duplex, asynchronous
2656 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2657 way, but not yet received, when the check is performed.
2659 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2660 on TCP/IP sockets, because this apparently causes some broken clients to
2663 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2664 unchanged) from the Cygwin maintainer.
2666 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2667 was supported instead of testing for mbx. Effectively a typo.
2669 15. The spa authenticator server code was not checking that the input it
2670 received was valid base64.
2672 16. The debug output line for the "set" modifier in ACLs was not showing the
2673 name of the variable that was being set.
2675 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2677 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2678 lookup on its own. Something else was needed to trigger the lookup. For
2679 example, a match in host_lookup or the need for a host name in a host list.
2680 Now, if $sender_host_name is referenced and the host name has not yet been
2681 looked up, a lookup is performed. If the lookup fails, the variable remains
2682 empty, and $host_lookup_failed is set to "1".
2684 19. Added "eqi" as a case-independent comparison operator.
2686 20. The saslauthd authentication condition could segfault if neither service
2687 nor realm was specified.
2689 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2690 error message that was logged was misleading, and incoming SMTP
2691 connections were dropped. The message is now more accurate, and temporary
2692 errors are given to SMTP connections.
2694 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2695 (except RSET) with a 421 error, until QUIT is received. However, it was
2696 failing to send a response to QUIT.
2698 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2699 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2700 ACL is now run after the helo_try_verify_hosts code.
2702 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2703 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2704 case-independent, so other case variants are also recognized). Apparently
2705 some systems use these upper case variants.
2707 25. If more than two messages were waiting for the same host, and a transport
2708 filter was specified for the transport, Exim sent two messages over the
2709 same TCP/IP connection, and then failed with "socket operation on non-
2710 socket" when it tried to send the third.
2712 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2714 27. The extern definition of crypt16() in expand.c was not being excluded when
2715 the OS had its own crypt16() function.
2717 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2718 as a preferred synonym for return_size_limit, both as an option and as an
2721 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2723 30. Changed the default configuration ACL to relax the local part checking rule
2724 for addresses that are not in any local domains. For these addresses,
2725 slashes and pipe symbols are allowed within local parts, but the sequence
2726 /../ is explicitly forbidden.
2728 31. SPA server authentication was not clearing the challenge buffer before
2731 32. log_message in a "warn" ACL statement was writing to the reject log as
2732 well as to the main log, which contradicts the documentation and doesn't
2733 seem right (because no rejection is happening). So I have stopped it.
2735 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2736 However, I am unable to do any testing of this.
2738 34. Fixed an infelicity in the appendfile transport. When checking directories
2739 for a mailbox, to see if any needed to be created, it was accidentally
2740 using path names with one or more superfluous leading slashes; tracing
2741 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2743 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2744 contents are added to the log line that is written for every discarded
2745 recipient. (Previously a log_message setting was ignored.)
2747 36. The ${quote: operator now quotes the string if it is empty.
2749 37. The install script runs exim in order to find its version number. If for
2750 some reason other than non-existence or emptiness, which it checks, it
2751 could not run './exim', it was installing it with an empty version number,
2752 i.e. as "exim-". This error state is now caught, and the installation is
2755 38. An argument was missing from the function that creates an error message
2756 when Exim fails to connect to the socket for saslauthd authentication.
2757 This could cause Exim to crash, or give a corrupted message.
2759 39. Added isip, isip4, and isip6 to ${if conditions.
2761 40. The ACL variables $acl_xx are now saved with the message, and can be
2762 accessed later in routers, transports, and filters.
2764 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2765 strings in the file are not string-expanded.
2767 42. If a MAIL command specified a SIZE value that was too large to fit into an
2768 int variable, the check against message_size_limit failed. Such values are
2769 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2770 day this will have to be increased, but I don't think I want to be around
2771 when emails are that large.
2778 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2779 iconv() is not standard in FreeBSD.
2781 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2782 IPv6 enabled. The observed symptom was a segmentation fault on return from
2783 the function os_common_find_running_interfaces() in src/os.c.
2785 3. In the check_special_case() function in daemon.c I had used "errno" as an
2786 argument name, which causes warnings on some systems. This was basically a
2787 typo, since it was named "eno" in the comments!
2789 4. The code that waits for the clock to tick (at a resolution of some fraction
2790 of a second) so as to ensure message-id uniqueness was always waiting for
2791 at least one whole tick, when it could have waited for less. [This is
2792 almost certainly not relevant at current processor speeds, where it is
2793 unlikely to ever wait at all. But we try to future-proof.]
2795 5. The function that sleeps for a time interval that includes fractions of a
2796 second contained a race. It did not block SIGALRM between setting the
2797 timer, and suspending (a couple of lines later). If the interval was short
2798 and the sigsuspend() was delayed until after it had expired, the suspension
2799 never ended. On busy systems this could lead to processes getting stuck for
2802 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2803 process, before it forks any delivery processes. The open lookup caching
2804 mechanism meant that the open file or database connection was passed into
2805 the delivery process. The problem was that delivery processes always tidy
2806 up cached lookup data. This could cause a problem for the next delivery
2807 process started by the queue runner, because the external queue runner
2808 process does not know about the closure. So the next delivery process
2809 still has data in the lookup cache. In the case of a file lookup, there was
2810 no problem because closing a file descriptor in a subprocess doesn't affect
2811 the parent. However, if the lookup was caching a connection to a database,
2812 the connection was closed, and the second delivery process was likely to
2813 see errors such as "PGSQL: query failed: server closed the connection
2814 unexpectedly". The problem has been fixed by closing all cached lookups
2815 in a queue runner before running a delivery process.
2817 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2818 seem to have the "const" qualifier which it has on other OS. I've
2821 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2822 *of the same type* (client or server) with the same public name that an
2823 error should be diagnosed.
2825 9. When Exim looked up a host name for an IP address, but failed to find the
2826 original IP address when looking up the host name (a safety check), it
2827 output the message "<ip address> does not match any IP for NULL", which was
2828 confusing, to say the least. The bug was that the host name should have
2829 appeared instead of "NULL".
2831 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2832 user that is built into the binary, and the -C or -D options is used, root
2833 privilege is dropped before the configuration file is read. In addition,
2834 logging is switched to stderr instead of the normal log files. If the
2835 configuration then re-defines the Exim user, the unprivileged environment
2836 is probably not what is expected, so Exim logs a panic warning message (but
2839 However, if deliver_drop_privilege is set, the unprivileged state may well
2840 be exactly what is intended, so the warning has been cut out in that case,
2841 and Exim is allowed to try to write to its normal log files.
2847 1. smtp_return_error_details was not giving details for temporary sender
2848 or receiver verification errors.
2850 2. Diagnose a configuration error if two authenticators have the same public
2853 3. Exim used not to create the message log file for a message until the first
2854 delivery attempt. This could be confusing when incoming messages were held
2855 for policy or load reasons. The message log file is now created at the time
2856 the message is received, and an initial "Received" line is written to it.
2858 4. The automatically generated man page for command line options had a minor
2859 bug that caused no ill effects; however, a more serious problem was that
2860 the procedure for building the man page automatically didn't always
2861 operate. Consequently, release 4.20 contains an out-of-date version. This
2862 shouldn't happen again.
2864 5. When building Exim with embedded Perl support, the script that builds the
2865 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2866 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2868 6. The freeze_tell option was not being used for messages that were frozen on
2869 arrival, either by an ACL or by local_scan().
2871 7. Added the smtp_incomplete_transaction log selector.
2873 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2874 was accepting AUTH without a new EHLO.
2876 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2877 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2880 10. Logging of TCP/IP connections (when configured) now happens in the main
2881 daemon process instead of the child process, so that the TCP/IP connection
2882 count is more accurate (but it can never be perfect).
2884 11. The use of "drop" in a nested ACL was not being handled correctly in the
2885 outer ACL. Now, if condition failure induced by the nested "drop" causes
2886 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2887 or "require"), the connection is dropped.
2889 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2890 that yield "discard" can now be used with an "accept" or a "discard" verb,
2891 but an error is generated for any others (because I can't see a useful way
2892 to define what should happen).
2894 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2895 are now processed in the same way as lines in the Exim configuration file.
2896 In particular, continuation lines are supported.
2898 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2900 15. Added -ti meaning -t -i.
2902 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2903 domains, and warn by logging if others are found.
2905 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2906 returned from GIFCONF when Exim is trying to find the list of interfaces on
2907 a host. The code in os.c has been modified to copy each ifreq to an aligned
2908 structure in all cases.
2910 Also, in some cases, the returned ifreq's were being copied to a 'struct
2911 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2912 means the last couple of bytes of an IPv6 address could be chopped if the
2913 ifreq contained only a normal sockaddr (14 bytes storage).
2915 18. Named domain lists were not supported in the hosts_treat_as_local option.
2916 An entry such as +xxxx was not recognized, and was treated as a literal
2919 19. Ensure that header lines added by a DATA ACL are included in the reject log
2920 if the ACL subsequently rejects the message.
2922 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2923 MD5 (which is deprecated).
2925 21. When testing a filter file using -bf, Exim was writing a message when it
2926 took the sender from a "From " line in the message, but it was not doing so
2927 when it took $return_path from a Return-Path: header line. It now does.
2929 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2930 with a valid header line field name (a series of printing characters
2931 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2933 23. Changed "disc" in the source to "disk" to conform to the documentation and
2934 the book and for uniformity.
2936 24. Ignore Sendmail's -Ooption=value command line item.
2938 25. When execve() failed while trying to run a command in a pipe transport,
2939 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2940 could be confused with a return value of 69 from the command itself. This
2941 has been changed to 127, the value the shell returns if it is asked to run
2942 a non-existent command. The wording for the related log line suggests a
2943 non-existent command as the problem.
2945 26. If received_header_text expands to an empty string, do not add a Received:
2946 header line to the message. (Well, it adds a token one on the spool, but
2947 marks it "old" so that it doesn't get used or transmitted.)
2949 27. Installed eximstats 1.28 (addition of -nt option).
2951 28. There was no check for failure on the call to getsockname() in the daemon
2952 code. This can fail if there is a shortage of resources on the system, with
2953 ENOMEM, for example. A temporary error is now given on failure.
2955 29. Contrary to the C standard, it seems that in some environments, the
2956 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2957 program. Exim now does this explicitly; it affects the formatting of
2958 timestamps using strftime().
2960 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2961 complaints. I've now initialized all the variables, to avoid this.
2963 32. Header lines added by a system filter were not being "seen" during
2964 transport-time rewrites.
2966 33. The info_callback() function passed to OpenSSL is set up with type void
2967 (*)(SSL *, int, int), as described somewhere. However, when calling the
2968 function (actually a macro) that sets it up, the type void(*)() is
2969 expected. I've put in a cast to prevent warnings from picky compilers.
2971 34. If a DNS black list lookup found a CNAME record, but there were no A
2972 records associated with the domain it pointed at, Exim crashed.
2974 35. If a DNS black list lookup returned more than one A record, Exim ignored
2975 all but the first. It now scans all returned addresses if a particular IP
2976 value is being sought. In this situation, the contents of the
2977 $dnslist_value variable are a list of all the addresses, separated by a
2980 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2981 to accept a host name and all its aliases if the forward lookup for any of
2982 them yielded the IP address of the incoming connection. Now it accepts only
2983 those names whose forward lookup yields the correct IP address. Any other
2984 names are discarded. This closes a loophole whereby a rogue DNS
2985 administrator could create reverse DNS records to break through a
2986 wildcarded host restriction in an ACL.
2988 37. If a user filter or a system filter that ran in a subprocess used any of
2989 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2990 the wrong values were passed to the pipe command ($thisaddress had the
2991 value of $0, $0 had the value of $1, etc). This bug was introduced by
2992 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2994 38. Improved the line breaking for long SMTP error messages from ACLs.
2995 Previously, if there was no break point between 40 and 75 characters, Exim
2996 left the rest of the message alone. Two changes have been made: (a) I've
2997 reduced the minimum length to 35 characters; (b) if it can't find a break
2998 point between 35 and 75 characters, it looks ahead and uses the first one
2999 that it finds. This may give the occasional overlong line, but at least the
3000 remaining text gets split now.
3002 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3003 file descriptors might be low, and that setting 1000 would always raise it.
3004 It turns out that in some environments, the limit is already over 1000 and
3005 that lowering it causes trouble. So now Exim takes care not to decrease it.
3007 40. When delivering a message, the value of $return_path is set to $sender_
3008 address at the start of routing (routers may change the value). By an
3009 oversight, this default was not being set up when an address was tested by
3010 -bt or -bv, which affected the outcome if any router or filter referred to
3013 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3014 log only when "message" or "log_message" are set. However, if one of the
3015 conditions was an address verification, or a call to a nested ACL, the
3016 messages generated by the underlying test were being passed through. This
3017 no longer happens. The underlying message is available in $acl_verify_
3018 message for both "message" and "log_message" expansions, so it can be
3019 passed through if needed.
3021 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3022 new expansion $bh_ to give the encoded byte string without charset
3023 translation. Translation happens only if iconv() is available; HAVE_ICONV
3024 indicates this at build time. HEADERS_CHARSET gives the charset to
3025 translate to; headers_charset can change it in the configuration, and
3026 "headers charset" can change it in an individual filter file.
3028 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3029 that creates RFC 2047 encoded "words" labels them as that charset instead
3030 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3031 expansion operator; (ii) when Exim creates a From: line for a local
3032 message; (iii) when a header line is rewritten to include a "phrase" part.
3034 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3035 buggy, causing it to skip the first lines of messages whose message ID
3036 ended in 'D'. This would not have bitten before Exim release 4.14, because
3037 message IDs were unlikely to end in 'D' before then. The effect was to have
3038 incorrect size information for certain domains.
3040 45. #include "config.h" was missing at the start of the crypt16.c module. This
3041 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3044 46. If there was a timeout during a "random" callout check, Exim treated it as
3045 a failure of the random address, and carried on sending RSET and the real
3046 address. If the delay was just some slowness somewhere, the response to the
3047 original RCPT would be taken as a response to RSET and so on, causing
3048 mayhem of various kinds.
3050 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3051 when I implemented it. It didn't allow for the fact that some option values
3052 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3053 the right test for positive values.
3055 48. Domain names in DNS records are case-independent. Exim always looks them up
3056 in lower case. Some resolvers return domain names in exactly the case they
3057 appear in the zone file, that is, they may contain uppercase letters. Not
3058 all resolvers do this - some return always lower case. Exim was treating a
3059 change of case by a resolver as a change of domain, similar to a widening
3060 of a domain abbreviation. This triggered its re-routing code and so it was
3061 trying to route what was effectively the same domain again. This normally
3062 caused routing to fail (because the router wouldn't handle the domain
3063 twice). Now Exim checks for this case specially, and just changes the
3064 casing of the domain that it ultimately uses when it transmits the message
3067 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3070 50. If a filter generated a file delivery with a non-absolute name (possible if
3071 no home directory exists for the router), the forbid_file option was not
3074 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3075 the existing equality matching.
3077 52. Exim was using ints instead of ino_t variables in some places where it was
3078 dealing with inode numbers.
3080 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3081 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3082 called TMPDIR, and if it finds it is different, it changes its value.
3084 54. The smtp_printf() function is now made available to local_scan() so
3085 additional output lines can be written before returning. There is also an
3086 smtp_fflush() function to enable the detection of a dropped connection.
3087 The variables smtp_input and smtp_batched_input are exported to
3090 55. Changed the default runtime configuration: the message "Unknown user"
3091 has been removed from the ACL, and instead placed on the localuser router,
3092 using the cannot_route_message feature. This means that any verification
3093 failures that generate their own messages won't get overridden. Similarly,
3094 the "Unrouteable address" message that was in the ACL for unverifiable
3095 relay addresses has also been removed.
3097 56. Added hosts_avoid_esmtp to the smtp transport.
3099 57. The exicyclog script was not checking for the esoteric option
3100 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3101 will work only if exicyclog is run under the appropriate euid.
3103 58. Following a discussion on the list, the rules by which Exim recognises line
3104 endings on incoming messages have been changed. The -dropcr and drop_cr
3105 options are now no-ops, retained only for backwards compatibility. The
3106 following line terminators are recognized: LF CRLF CR. However, special
3107 processing applies to CR:
3109 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3110 nor a local message in the state where . is a terminator.
3112 (ii) If a bare CR is encountered in a header line, an extra space is added
3113 after the line terminator so as not to end the header. The reasoning
3114 behind this is that bare CRs in header lines are most likely either
3115 to be mistakes, or people trying to play silly games.
3117 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3118 was being incorrectly given as 18 bytes larger than it should have been.
3119 This is a VOB (very old bug).
3121 60. This may never have affected anything current, but just in case it has:
3122 When the local host is found other than at the start of a list of hosts,
3123 the local host, those with the same MX, and any that follow, are discarded.
3124 When the list in question was part of a longer list of hosts, the following
3125 hosts (not currently being processed) were also being discarded. This no
3126 longer happens. I'm not sure if this situation could ever has previously
3129 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3132 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3133 header. This is something that is recommended in a new Internet Draft, and
3134 is something that is documented as being done by Sendmail. There are two
3135 possible values. For messages generated by the autoreply transport, Exim
3138 Auto-Submitted: auto-replied
3140 whereas for all other generated messages (e.g. bounces) it adds
3142 Auto-Submitted: auto-generated
3144 63. The "personal" condition in filters now includes a test for the
3145 Auto-Submitted: header. If it contains the string "auto-" the message it
3146 not considered personal.
3148 64. Added rcpt_include_affixes as a generic transport option.
3150 65. Added queue_only_override (default true).
3152 66. Added the syslog_duplication option.
3154 67. If what should have been the first header line of a message consisted of
3155 a space followed by a colon, Exim was mis-interpreting it as a header line.
3156 It isn't of course - it is syntactically invalid and should therefore be
3157 treated as the start of the message body. The misbehaviour could have
3158 caused a number of strange effects, including loss of data in subsequent
3159 header lines, and spool format errors.
3161 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3162 client host had authenticated. This control can now be exercised by an ACL
3163 for more flexibility.
3165 69. By default, callouts do not happen when testing with -bh. There is now a
3166 variant, -bhc, which does actually run the callout code, including
3167 consulting and updating the callout cache.
3169 70. Added support for saslauthd authentication, courtesy of Alexander
3172 71. If statvfs() failed on the spool or log directories while checking their
3173 size for availability, Exim confusingly gave the error "space shortage".
3174 Furthermore, in debugging mode it crashed with a floating point exception.
3175 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3176 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3177 serious problem, Exim now writes to the main and panic logs when this
3178 happens, with details of the failure. It then refuses to accept the
3179 incoming message, giving the message "spool directory problem" or "log
3180 directory problem" with a 421 code for SMTP messages.
3182 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3183 0, 1, and 2 exist, because some OS complain for execs without them (see
3184 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3185 descriptors. However, the code omitted to check that the open succeeded,
3186 causing mysterious errors if for some reason the permissions on /dev/null
3187 got screwed. Now Exim writes a message to the main and panic logs, and
3188 bombs out if it can't open /dev/null.
3190 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3191 interact so that it is all more flexible. It is supposed to remain
3192 backwards compatible. Also added extra_local_interfaces.
3194 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3195 to bomb out with an assertion failure - to the client this appears as a
3196 connection drop. This problem occurs in the part of the code that was taken
3197 from the Samba project. Fortunately, the assertion is in a very simple
3198 function, so I have fixed this by reproducing the function inline in the
3199 one place where it is called, and arranging for authentication to fail
3200 instead of killing the process with assert().
3202 75. The SPA client code was not working when the server requested OEM rather
3203 than Unicode encoding.
3205 76. Added code to make require_files with a specific uid setting more usable in
3206 the case where statting the file as root fails - usually a non-root-mounted
3207 NFS file system. When this happens and the failure is EACCES, Exim now
3208 forks a subprocess and does the per-uid checking as the relevant uid.
3210 77. Added process_log_path.
3212 78. If log_file_path was not explicitly set, a setting of check_log_space or
3213 check_log_inodes was ignored.
3215 79. If a space check for the spool or log partitions fails, the incident is now
3216 logged. Of course, in the latter case the data may get lost...
3218 80. Added the %p formatting code to string_format() so that it can be used to
3219 print addresses in debug_print(). Adjusted all the address printing in the
3220 debugging in store.c to use %p rather than %d.
3222 81. There was a concern that a line of code in smtp_in.c could overflow a
3223 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3224 initially expressed, the concern was not well-founded, because trailing
3225 spaces are removed early. However, if the trailing spaces were followed by
3226 a NULL, they did not get removed, so the overflow was possible. Two fixes
3229 (a) I re-wrote the offending code in a cleaner fashion.
3230 (b) If an incoming SMTP command contains a NULL character, it is rejected
3233 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3234 runs initgroups(), so that if the Exim user is in any additional groups,
3235 they will be used during message reception.
3241 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob