1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.200 2005/08/08 10:29:08 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.
74 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
76 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
77 command itself ended in error, the underlying message about the transport
78 filter timeout was being overwritten with the pipe command error. Now the
79 underlying error message should be appended to the second error message.
81 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
83 PH/02 Reduce the amount of output that "make" produces by default. Full output
84 can still be requested.
86 PH/03 The warning log line about a condition test deferring for a "warn" verb
87 was being output only once per connection, rather than after each
88 occurrence (because it was using the same function as for successful
89 "warn" verbs). This seems wrong, so I have changed it.
91 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
92 it should not have, which might have caused a crash in the right
93 circumstances, but probably never did.
95 PH/04 Installed a modified version of Tony Finch's patch to make submission
96 mode fix the return path as well as the Sender: header line, and to
97 add a /name= option so that you can make the user's friendly name appear
100 TF/03 Added the control = fakedefer ACL modifier.
102 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
103 Mark Lowes for thorough testing.
105 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
107 TK/03 Merged latest SRS patch from Miles Wilton.
109 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
110 with the definition in sysexits.h (which is #included earlier).
111 Fortunately, Exim does not actually use EX_OK. The code used to try to
112 preserve the sysexits.h value, by assumimg that macro definitions were
113 scanned for macro replacements. I have been disabused of this notion,
114 so now the code just undefines EX_OK before #including unistd.h.
116 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
117 in the smtp transport. When a block could not be written in a single
118 write() function, the timeout was being re-applied to each part-write.
119 This seems wrong - if the receiver was accepting one byte at a time it
120 would take for ever. The timeout is now adjusted when this happens. It
121 doesn't have to be particularly precise.
123 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
124 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
126 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
127 without a domain if the check to <postmaster@domain> fails.
129 SC/01 Eximstats: added -xls and the ability to specify output files
130 (patch written by Frank Heydlauf).
132 SC/02 Eximstats: use FileHandles for outputing results.
134 SC/03 Eximstats: allow any combination of xls, txt, and html output.
136 SC/04 Eximstats: fixed display of large numbers with -nvr option
138 SC/05 Eximstats: fixed merging of reports with empty tables.
140 SC/06 Eximstats: added the -include_original_destination flag
142 SC/07 Eximstats: removed tabs and trailing whitespace.
144 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
146 TK/06 MBOX spool code: Add real "From " MBOX separator line
147 so the .eml file is really in mbox format (even though
148 most programs do not really care). Patch from Alex Miller.
150 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
151 The latter is generated from $received_to and is only set if the
152 message has one envelope recipient. SA can use these headers,
153 obviously out-of-the-box. Patch from Alex Miller.
155 PH/08 The ${def test on a variable was returning false if the variable's
156 value was "0", contrary to what the specification has always said!
157 The result should be true unless the variable is empty.
159 PH/09 The syntax error of a character other than { following "${if
160 def:variable_name" (after optional whitespace) was not being diagnosed.
161 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
162 accidental colon was present, for example, could give incorrect results.
164 PH/10 Tidied the code in a number of places where the st_size field of a stat()
165 result is used (not including appendfile, where other changes are about
168 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
169 This involved changing a lot of size variables from int to off_t. It
170 should work with maildirs and everything.
172 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
173 spamd dying while we are connected to it.
175 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
178 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
179 function for BDB 4.3.
181 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
183 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
184 the sources for PCRE have been re-arranged and more files are now
187 PH/15 The code I had for printing potentially long long variables in PH/11
188 above was not the best (it lost precision). The length of off_t variables
189 is now inspected at build time, and an appropriate printing format (%ld
190 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
191 to be "long long int" or "long int". This is needed for the internal
192 formatting function string_vformat().
194 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
195 the configuration file to be ":syslog", then the script "guesses" where
196 the logs files are, rather than using the compiled in default. In our
197 case the guess is not the same as the compiled default, so the script
198 suddenly stopped working when I started to use syslog. The patch checks
199 to see if log_file_path is "". If so, it attempts to read it from exim
200 with no configuration file to get the compiled in version, before it
201 falls back to the previous guessing code."
203 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
204 implementing BATV in an Exim configuration. See NewStuff for the gory
207 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
208 Makefile that are specific to HP-UX.
210 PH/18 If the "use_postmaster" option was set for a recipient callout together
211 with the "random" option, the postmaster address was used as the MAIL
212 FROM address for the random test, but not for the subsequent recipient
213 test. It is now used for both.
215 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
216 patch removes a few documentation additions to RFC 3028, because the
217 latest draft now contains them. It adds the new en;ascii-case comparator
218 and a new error check for 8bit text in MIME parts. Comparator and
219 require names are now matched exactly. I enabled the subaddress
220 extension, but it is not well tested yet (read: it works for me)."
222 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
223 rework some of the code of TK/09 above to avoid the hardwired use of
224 "%lld" and "long long". Replaced the call to snprintf() with a call to
227 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
228 records point to non-existent hosts", "retry timeout exceeded", and
229 "retry time not reached for any host after a long failure period".
231 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
232 experimental DomainKeys support:
234 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
235 (2) On an error such as an illegally used "control", the wrong name for
236 the control was given.
238 These problems did NOT occur unless DomainKeys support was compiled.
240 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
242 PH/24 Added ${if match_ip condition.
244 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
245 databases so that it will be absolutely obvious if a crash occurs in the
246 DB library. This is a regular occurrence (often caused by mis-matched
249 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
250 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
251 picked up on a user's system that detects such things. There doesn't seem
252 to be a gcc warning option for this - only an attribute that has to be
253 put on the function's prototype. It seems that in Fedora Core 4 they have
254 set this on a number of new functions. No doubt there will be more in due
257 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
258 specify a transport. However, if an address that was verified by such a
259 router was the subject of a callout, Exim crashed because it tried to
260 read the rcpt_include_affixes from the non-existent transport. Now it
261 just assumes that the setting of that option is false. This bug was
262 introduced by 4.51/PH/31.
264 PH/28 Changed -d+all to exclude +memory, because that information is very
265 rarely of interest, but it makes the output a lot bigger. People tend to
266 do -d+all out of habit.
268 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
269 code in os-type was giving problems when libc.so lives in lib64, like on
272 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
273 aren't the modern standard, and it seems that some systems' include files
274 don't always have them. Exim was already checking for some of the newer
275 ones like T_AAAA, and defining it itself. I've added checks for all the
276 record types that Exim uses.
278 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
279 not automatically generating a new one, as it is supposed to. This
280 prevented TLS from working. If the file did exist, but contained invalid
281 data, a new version was generated, as expected. It was only the case of a
282 non-existent file that was broken.
284 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
285 with a change in libdomainkeys > 0.64.
287 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
288 from DNS. If the selector record carries the flag, it now has
289 precedence over the domain-wide flag.
291 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
293 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
294 the use of a port name that isn't defined in /etc/services) occurred, the
295 message was deferred as in a normal delivery, and thus remained on the
296 spool, instead of being failed because of the mua_wrapper setting. This
297 is now fixed, and I tidied up some of the mua_wrapper messages at the
300 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
301 the messages in a hash of arrays rather than using individual hashes.
302 This is a bit cleaner and results in dramatic memory savings, albeit
303 at a slight CPU cost.
305 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
306 as requested by Marc Sherman.
308 SC/10 Eximstats: added histograms for user specified patterns as requested
311 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
313 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
314 fopen() in the content-scanning modules that did not already have it.
320 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
321 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
323 TK/02 Fix ACL "control" statement not being available in MIME ACL.
325 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
327 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
328 to test Sieve filters that use "vacation".
330 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
331 that changes the way the GnuTLS parameters are stored in the cache file.
332 The new format can be generated externally. For backward compatibility,
333 if the data in the cache doesn't make sense, Exim assumes it has read an
334 old-format file, and it generates new data and writes a new file. This
335 means that you can't go back to an older release without removing the
338 PH/03 A redirect router that has both "unseen" and "one_time" set does not
339 work if there are any delivery delays because "one_time" forces the
340 parent to be marked "delivered", so its unseen clone is never tried
341 again. For this reason, Exim now forbids the simultaneous setting of
344 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
345 redirected to themselves ("homonym" addresses). Read the long ChangeLog
346 entry if you want to know the details. The fix, however, neglected to
347 consider the case when local delivery batching is involved. The test for
348 "previously delivered" was not happening when checking to see if an
349 address could be batched with a previous (undelivered) one; under
350 certain circumstances this could lead to multiple deliveries to the same
353 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
354 in its include files, and this causes problems building Exim.
356 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
357 header_syntax) but Exim was just ignoring anything given after a slash.
358 In particular, this caused confusion with an attempt to use "verify =
359 reverse_host_lookup/defer_ok". An error is now given when options are
360 supplied for verify items that do not have them. (Maybe reverse_host_
361 lookup should have a defer_ok option, but that's a different point.)
363 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
364 defined by RFC 821) to 2048, because there were problems with some AUTH
365 commands, and RFC 1869 says the size should be increased for extended
366 SMTP commands that take arguments.
368 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
371 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
372 "unknown" error; now it says that the functionality isn't in the binary.
374 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
375 an address' error message when a string expansion fails (syntax or
376 whatever). Otherwise the password may appear in the log. Following change
377 PH/42 below, there is no longer a chance of it appearing in a bounce
380 PH/11 Installed exipick version 20050225.0 from John Jetmore.
382 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
383 of its addresses was ever tried. (Bugzilla bug #2.)
385 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
386 the result incorrectly in the debug output. (It correctly added a newline
387 to what was transported.)
389 TF/01 Added $received_time.
391 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
392 commented out examples of how to interface to a virus scanner and to
393 SpamAssassin. Also added commented examples of av_scanner and
394 spamd_address settings.
396 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
397 and controls are allowed in which ACLs. There were a couple of minor
398 errors. Some of the entries in the conditions table (which is a table of
399 where they are NOT allowed) were getting very unwieldy; rewrote them as a
400 negation of where the condition IS allowed.
402 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
404 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
405 header file does not have a version number, so I've had to invent a new
406 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
407 API. The code is untested by me (my Linux distribution still has 0.3.2 of
408 radiusclient), but it was contributed by a Radius user.
410 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
411 files or directories, for OpenSSL.
413 PH/19 When an Exim process that is running as root has to create an Exim log
414 file, it does so in a subprocess that runs as exim:exim so as to get the
415 ownership right at creation (otherwise, other Exim processes might see
416 the file with the wrong ownership). There was no test for failure of this
417 fork() call, which would lead to the process getting stuck as it waited
418 for a non-existent subprocess. Forks do occasionally fail when resources
419 run out. I reviewed all the other calls to fork(); they all seem to check
422 PH/20 When checking for unexpected SMTP input at connect time (before writing
423 the banner), Exim was not dealing correctly with a non-positive return
424 from the read() function. If the client had disconnected by this time,
425 the result was a log entry for a synchronization error with an empty
426 string after "input=" when read() returned zero. If read() returned -1
427 (an event I could not check), uninitialized data bytes were printed.
428 There were reports of junk text (parts of files, etc) appearing after
431 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
433 PH/22 Added support for macro redefinition, and (re)definition in between
434 driver and ACL definitions.
436 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
437 forgetting to use the resulting value; it was using the unexpanded value.
439 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
440 hadn't been configured. The fix is from Juergen Kreileder, who
441 understands it better than I do:
443 "Here's what I see happening with three configured cyrus_sasl
444 authenticators configured (plain, login, cram-md5):
446 On startup auth_cyrus_sasl_init() gets called for each of these.
447 This means three calls to sasl_listmech() without a specified mech_list.
448 => SASL tests which mechs of all available mechs actually work
449 => three warnings about OTP not working
450 => the returned list contains: plain, login, cram-md5, digest-md5, ...
452 With the patch, sasl_listmech() also gets called three times. But now
453 SASL's mech_list option is set to the server_mech specified in the the
454 authenticator. Or in other words, the answer from sasl_listmech()
455 gets limited to just the mech you're testing for (which is different
457 => the return list contains just 'plain' or 'login', 'cram-md5' or
458 nothing depending on the value of ob->server_mech.
460 I've just tested the patch: Authentication still works fine,
461 unavailable mechs specified in the exim configuration are still
462 caught, and the auth.log warnings about OTP are gone."
464 PH/25 When debugging is enabled, the contents of the command line are added
465 to the debugging output, even when log_selector=+arguments is not
468 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
469 answer is "GNU", and only if the return is "GNU/something" is the answer
472 PH/27 $acl_verify_message is now set immediately after the failure of a
473 verification in an ACL, and so is available in subsequent modifiers. In
474 particular, the message can be preserved by coding like this:
476 warn !verify = sender
477 set acl_m0 = $acl_verify_message
479 Previously, $acl_verify_message was set only while expanding "message"
480 and "log_message" when a very denied access.
482 PH/28 Modified OS/os.c-Linux with
484 -#ifndef OS_LOAD_AVERAGE
485 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
487 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
488 nomenclature these days.)
490 PH/29 Installed patch from the Sieve maintainer that adds the options
491 sieve_useraddress and sieve_subaddress to the redirect router.
493 PH/30 In these circumstances:
494 . Two addresses routed to the same list of hosts;
495 . First host does not offer TLS;
496 . First host accepts first address;
497 . First host gives temporary error to second address;
498 . Second host offers TLS and a TLS session is established;
499 . Second host accepts second address.
500 Exim incorrectly logged both deliveries with the TLS parameters (cipher
501 and peerdn, if requested) that were in fact used only for the second
504 PH/31 When doing a callout as part of verifying an address, Exim was not paying
505 attention to any local part prefix or suffix that was matched by the
506 router that accepted the address. It now behaves in the same way as it
507 does for delivery: the affixes are removed from the local part unless
508 rcpt_include_affixes is set on the transport.
510 PH/32 Add the sender address, as F=<...>, to the log line when logging a
511 timeout during the DATA phase of an incoming message.
513 PH/33 Sieve envelope tests were broken for match types other than :is. I have
514 applied a patch sanctioned by the Sieve maintainer.
516 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
517 the uid or gid is negative. A case of a negative gid caused this to be
518 noticed. The fix allows for either to be negative.
520 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
521 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
524 PH/36 The change PH/12 above was broken. Fixed it.
526 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
527 the grounds that routing the same address twice would always produce the
528 same answer. This might have been true once, but it is certainly no
529 longer true now. Routing a child address may depend on the previous
530 routing that produced that child. Some complicated redirection strategies
531 went wrong when messages had multiple recipients, and made Exim's
532 behaviour dependent on the order in which the addresses were given.
534 I have moved the duplicate checking until after the routing is complete.
535 Exim scans the addresses that are assigned to local and remote
536 transports, and removes any duplicates. This means that more work will be
537 done, as duplicates will always all be routed, but duplicates are
538 presumably rare, so I don't expect this is of any significance.
540 For deliveries to pipes, files, and autoreplies, the duplicate checking
541 still happens during the routing process, since they are not going to be
544 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
545 It corrects a timeout issue with spamd. This is Ian's comment: "The
546 background is that sometimes spamd either never reads data from a
547 connection it has accepted, or it never writes response data. The exiscan
548 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
549 blindly assumes that writes won't block so it may never time out."
551 PH/39 Allow G after quota size as well as K and M.
553 PH/40 The value set for $authenticated_id in an authenticator may not contain
554 binary zeroes or newlines because the value is written to log lines and
555 to spool files. There was no check on this. Now the value is run through
556 the string_printing() function so that such characters are converted to
557 printable escape sequences.
559 PH/41 $message_linecount is a new variable that contains the total number of
560 lines in the message. Compare $body_linecount, which is the count for the
563 PH/42 Exim no longer gives details of delivery errors for specific addresses in
564 bounce and delay warning messages, except in certain special cases, which
567 (a) An SMTP error message from a remote host;
568 (b) A message specified in a :fail: redirection;
569 (c) A message specified in a "fail" command in a system filter;
570 (d) A message specified in a FAIL return from the queryprogram router;
571 (e) A message specified by the cannot_route_message router option.
573 In these cases only, Exim does include the error details in bounce and
574 warning messages. There are also a few cases where bland messages such
575 as "unrouteable address" or "local delivery error" are given.
577 PH/43 $value is now also set for the "else" part of a ${run expansion.
579 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
580 being worked on, but at least Exim now implements the latest version to
583 PH/45 In a pipe transport, although a timeout while waiting for the pipe
584 process to complete was treated as a delivery failure, a timeout while
585 writing the message to the pipe was logged, but erroneously treated as a
586 successful delivery. Such timeouts include transport filter timeouts. For
587 consistency with the overall process timeout, these timeouts are now
588 treated as errors, giving rise to delivery failures by default. However,
589 there is now a new Boolean option for the pipe transport called
590 timeout_defer, which, if set TRUE, converts the failures into defers for
591 both kinds of timeout. A transport filter timeout is now identified in
594 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
595 systems where "make" and "gmake" are different, calling "gmake" at top
596 level broke things. I've arranged for the value of $(MAKE) to be passed
597 from the Makefile to this script so that it can call the same version of
601 A note about Exim versions 4.44 and 4.50
602 ----------------------------------------
604 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
605 changes of various kinds. As a consequence, a big documentation update was
606 needed. This delayed the release for rather longer than seemed good, especially
607 in the light of a couple of (minor) security issues. Therefore, the changes
608 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
609 release. So 4.44 and 4.50 are in effect two different branches that both start
612 I have left the 4.50 change log unchanged; it contains all the changes since
613 4.43. The change log for 4.44 is below; many of its items are identical to
614 those for 4.50. This seems to be the most sensible way to preserve the
615 historical information.
621 1. Minor wording change to the doc/README.SIEVE file.
623 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
624 computation of the current number of files was incorrect.
626 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
627 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
630 4. Give more explanation in the error message when the command for a transport
631 filter fails to execute.
633 5. There are several places where Exim runs a non-Exim command in a
634 subprocess. The SIGUSR1 signal should be disabled for these processes. This
635 was being done only for the command run by the queryprogram router. It is
636 now done for all such subprocesses. The other cases are: ${run, transport
637 filters, and the commands run by the lmtp and pipe transports.
639 6. Added CONFIGURE_GROUP build-time option.
641 7. Some older OS have a limit of 256 on the maximum number of file
642 descriptors. Exim was using setrlimit() to set 1000 as a large value
643 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
644 systems. I've change it so that if it can't get 1000, it tries for 256.
646 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
647 was an oversight, and furthermore, ever since the addition of extra
648 controls (e.g. 4.43/32), the checks on when to allow different forms of
649 "control" were broken. There should now be diagnostics for all cases when a
650 control that does not make sense is encountered.
652 9. Added the /retain_sender option to "control=submission".
654 10. $recipients is now available in the predata ACL (oversight).
656 11. Tidy the search cache before the fork to do a delivery from a message
657 received from the command line. Otherwise the child will trigger a lookup
658 failure and thereby defer the delivery if it tries to use (for example) a
659 cached ldap connection that the parent has called unbind on.
661 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
662 of $address_data from the recipient verification was clobbered by the
665 13. The value of address_data from a sender verification is now available in
666 $sender_address_data in subsequent conditions in the ACL statement.
668 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
670 15. Added a new option "connect=<time>" to callout options, to set a different
673 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
674 was its contents. (It was OK if the option was not defined at all.)
676 17. A "Completed" log line is now written for messages that are removed from
677 the spool by the -Mrm option.
679 18. New variables $sender_verify_failure and $recipient_verify_failure contain
680 information about exactly what failed.
682 19. Added -dd to debug only the daemon process.
684 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
685 handles timeouts, both on the server side and network timeouts. Renamed the
686 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
688 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
689 It is now set to "smtps".
691 22. $host_address is now set to the target address during the checking of
694 23. When checking ignore_target_hosts for an ipliteral router, no host name was
695 being passed; this would have caused $sender_host_name to have been used if
696 matching the list had actually called for a host name (not very likely,
697 since this list is usually IP addresses). A host name is now passed as
700 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
701 code that specifies a non-restarting handler (typically sigaction() in
702 modern systems) in an attempt to fix a rare and obscure crash bug.
704 25. Narrowed the window for a race in the daemon that could cause it to ignore
705 SIGCHLD signals. This is not a major problem, because they are used only to
706 wake it up if nothing else does.
708 26. A malformed maildirsize file could cause Exim to calculate negative values
709 for the mailbox size or file count. Odd effects could occur as a result.
710 The maildirsize information is now recalculated if the size or filecount
713 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
714 support for a long time. Removed HAVE_SYS_VFS_H.
716 28. Installed the latest version of exipick from John Jetmore.
718 29. In an address list, if the pattern was not a regular expression, an empty
719 subject address (from a bounce message) matched only if the pattern was an
720 empty string. Non-empty patterns were not even tested. This was the wrong
721 because it is perfectly reasonable to use an empty address as part of a
722 database query. An empty address is now tested by patterns that are
723 lookups. However, all the other forms of pattern expect the subject to
724 contain a local part and a domain, and therefore, for them, an empty
725 address still always fails if the pattern is not itself empty.
727 30. Exim went into a mad DNS loop when attempting to do a callout where the
728 host was specified on an smtp transport, and looking it up yielded more
731 31. Re-factored the code for checking spool and log partition space into a
732 function that finds that data and another that does the check. The former
733 is then used to implement four new variables: $spool_space, $log_space,
734 $spool_inodes, and $log_inodes.
736 32. The RFC2047 encoding function was originally intended for short strings
737 such as real names; it was not keeping to the 75-character limit for
738 encoded words that the RFC imposes. It now respects the limit, and
739 generates multiple encoded words if necessary. To be on the safe side, I
740 have increased the buffer size for the ${rfc2047: expansion operator from
743 33. It is now permitted to omit both strings after an "if" condition; if the
744 condition is true, the result is "true". As before, when the second string
745 is omitted, a false condition yields an empty string. This makes it less
746 cumbersome to write custom ACL and router conditions.
748 34. Failure to deliver a bounce message always caused it to be frozen, even if
749 there was an errors_to setting on the router. The errors_to setting is now
752 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
753 canonical form (fully expanded) before being placed in
754 $sender_host_address.
756 36. The table in the code that translates DNS record types into text (T_A to
757 "A" for instance) was missing entries for NS and CNAME. It is just possible
758 that this could have caused confusion if both these types were looked up
759 for the same domain, because the text type is used as part of Exim's
760 per-process caching. But the chance of anyone hitting this buglet seems
763 37. The dnsdb lookup has been extended in a number of ways.
765 (1) There is a new type, "zns", which walks up the domain tree until it
766 finds some nameserver records. It should be used with care.
768 (2) There is a new type, "mxh", which is like "mx" except that it returns
769 just the host names, not the priorities.
771 (3) It is now possible to give a list of domains (or IP addresses) to be
772 looked up. The behaviour when one of the lookups defers can be
773 controlled by a keyword.
775 (4) It is now possible to specify the separator character for use when
776 multiple records are returned.
778 38. The dnslists ACL condition has been extended: it is now possible to supply
779 a list of IP addresses and/or domains to be looked up in a particular DNS
782 39. Added log_selector=+queue_time_overall.
784 40. When running the queue in the test harness, wait just a tad after forking a
785 delivery process, to get repeatability of debugging output.
787 41. Include certificate and key file names in error message when GnuTLS fails
788 to set them up, because the GnuTLS error message doesn't include the name
789 of the failing file when there is a problem reading it.
791 42. Allow both -bf and -bF in the same test run.
793 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
795 44. The "Exiscan patch" is now merged into the mainline Exim source.
797 45. Sometimes the final signoff response after QUIT could fail to get
798 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
799 < 0 before doing a fflush(). This bug looks as though it goes back to the
800 introduction of TLS in release 3.20, but "sometimes" must have been rare
801 because the tests only now provoked it.
803 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
804 (this can affect the format of dates).
806 47. exim_tidydb, when checking for the continued existence of a message for
807 which it has found a message-specific retry record, was not finding
808 messages that were in split spool directories. Consequently, it was
809 deleting retry records that should have stayed in existence.
811 48. Steve fixed some bugs in eximstats.
813 49. The SPA authentication driver was not abandoning authentication and moving
814 on to the next authenticator when an expansion was forced to fail,
815 contradicting the general specification for all authenticators. Instead it
816 was generating a temporary error. It now behaves as specified.
818 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
819 (the order specifies the preference for clients). The order is now AES256,
820 AES128, 3DES, ARCFOUR128.
822 51. Small patch to Sieve code - explicitly set From: when generating an
825 52. Exim crashed if a remote delivery caused a very long error message to be
826 recorded - for instance if somebody sent an entire SpamAssassin report back
827 as a large number of 550 error lines. This bug was coincidentally fixed by
828 increasing the size of one of Exim's internal buffers (big_buffer) that
829 happened as part of the Exiscan merge. However, to be on the safe side, I
830 have made the code more robust (and fixed the comments that describe what
833 53. Now that there can be additional text after "Completed" in log lines (if
834 the queue_time_overall log selector is set), a one-byte patch to exigrep
835 was needed to allow it to recognize "Completed" as not the last thing in
838 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
839 patch that reportedly fixes this has been added. I am not expert enough to
840 create a test for it. This is what the patch creator wrote:
842 "I found a little strange behaviour of ldap code when working with
843 Windows 2003 AD Domain, where users was placed in more than one
844 Organization Units. When I tried to give exim partial DN, the exit code
845 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
846 But simultaneously result of request was absolutely normal ldap result,
847 so I produce this patch..."
849 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
850 REFERENCE, so I have modified the code to exclude the patch when that macro
853 55. Some experimental protocols are using DNS PTR records for new purposes. The
854 keys for these records are domain names, not reversed IP addresses. The
855 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
856 leaves it alone. Component reversal etc. now happens only for IP addresses.
859 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
861 57. Double the size of the debug message buffer (to 2048) so that more of very
862 long debug lines gets shown.
864 58. The exicyclog utility now does better if the number of log files to keep
865 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
867 59. Two changes related to the smtp_active_hostname option:
869 (1) $smtp_active_hostname is now available as a variable.
870 (2) The default for smtp_banner uses $smtp_active_hostname instead
871 of $primary_hostname.
873 60. The host_aton() function is supposed to be passed a string that is known
874 to be a valid IP address. However, in the case of IPv6 addresses, it was
875 not checking this. This is a hostage to fortune. Exim now panics and dies
876 if the condition is not met. A case was found where this could be provoked
877 from a dnsdb PTR lookup with an IPv6 address that had more than 8
878 components; fortuitously, this particular loophole had already been fixed
879 by change 4.50/55 above.
881 If there are any other similar loopholes, the new check in host_aton()
882 itself should stop them being exploited. The report I received stated that
883 data on the command line could provoke the exploit when Exim was running as
884 exim, but did not say which command line option was involved. All I could
885 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
889 61. There was a buffer overflow vulnerability in the SPA authentication code
890 (which came originally from the Samba project). I have added a test to the
891 spa_base64_to_bits() function which I hope fixes it.
894 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
895 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
897 63. The daemon start-up calls getloadavg() while still root for those OS that
898 need the first call to be done as root, but it missed one case: when
899 deliver_queue_load_max is set with deliver_drop_privilege. This is
900 necessary for the benefit of the queue runner, because there is no re-exec
901 when deliver_drop_privilege is set.
903 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
906 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
907 was in use, was not putting the data itself into the right store pool;
908 consequently, it could be overwritten for a subsequent message in the same
909 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
912 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
914 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
915 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
916 the function were treating the return as a boolean value, which happened to
917 work because 0=false and not-0=true, but is not correct code.
919 68. The host_aton() function was not handling scoped IPv6 addresses (those
920 with, for example, "%eth0" on the end) correctly.
922 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
923 negated items (that is, ~something) in unsigned ints. Some compilers
924 apparently mutter when there is no cast.
926 70. If an address verification called from an ACL failed, and did not produce a
927 user-specific message (i.e. there was only a "system" message), nothing was
928 put in $acl_verify_message. In this situation, it now puts the system
931 71. Change 4.23/11 added synchronization checking at the start of an SMTP
932 session; change 4.31/43 added the unwanted input to the log line - except
933 that it did not do this in the start of session case. It now does.
935 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
936 This is wrong and can cause the other end to generate a synchronization
937 error if it is another Exim or anything else that does the synchronization
938 check. A QUIT command is no longer sent after a timeout.
940 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
943 74. The defer_ok option of callout verification was not working if it was used
944 when verifying addresses in header lines, that is, for this case:
946 verify = header_sender/callout=defer_ok
948 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
949 those file descriptors could be used for SMTP connections. If anything
950 wrote to stderr (the example that came up was "warn" in embedded Perl), it
951 could be sent to the SMTP client, causing chaos. The daemon now opens
952 stdin, stdout, and stderr to /dev/null when it puts itself into the
955 76. Arrange for output from Perl's "warn" command to be written to Exim's main
956 log by default. The user can override this with suitable Perl magic.
958 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
959 the log message when discard triggers, was not working for the DATA ACL or
960 for the non-SMTP ACL.
962 78. Error message wording change in sieve.c.
964 79. If smtp_accept_max_per_host was set, the number of connections could be
965 restricted to fewer than expected, because the daemon was trying to set up
966 a new connection before checking whether the processes handling previous
967 connections had finished. The check for completed processes is now done
968 earlier. On busy systems, this bug wouldn't be noticed because something
969 else would have woken the daemon, and it would have reaped the completed
972 80. If a message was submitted locally by a user whose login name contained one
973 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
974 It caused a spool format error. I have fixed the spool reading code. A
975 related problem was that the "from" clause in the Received: line became
976 illegal because of the space(s). It is now covered by ${quote_local_part.
978 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
981 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
983 83. Patch from Sieve maintainer to fix unterminated string problem in
986 84. Some minor changes to the Linux configuration files to help with other
987 OS variants using glibc.
989 85. One more patch for Sieve to update vacation handling to latest spec.
992 ----------------------------------------------------
993 See the note above about the 4.44 and 4.50 releases.
994 ----------------------------------------------------
1000 1. Change 4.43/35 introduced a bug that caused file counts to be
1001 incorrectly computed when quota_filecount was set in an appendfile
1004 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1005 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1008 3. Give more explanation in the error message when the command for a transport
1009 filter fails to execute.
1011 4. There are several places where Exim runs a non-Exim command in a
1012 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1013 was being done only for the command run by the queryprogram router. It is
1014 now done for all such subprocesses. The other cases are: ${run, transport
1015 filters, and the commands run by the lmtp and pipe transports.
1017 5. Some older OS have a limit of 256 on the maximum number of file
1018 descriptors. Exim was using setrlimit() to set 1000 as a large value
1019 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1020 systems. I've change it so that if it can't get 1000, it tries for 256.
1022 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1023 was an oversight, and furthermore, ever since the addition of extra
1024 controls (e.g. 4.43/32), the checks on when to allow different forms of
1025 "control" were broken. There should now be diagnostics for all cases when a
1026 control that does not make sense is encountered.
1028 7. $recipients is now available in the predata ACL (oversight).
1030 8. Tidy the search cache before the fork to do a delivery from a message
1031 received from the command line. Otherwise the child will trigger a lookup
1032 failure and thereby defer the delivery if it tries to use (for example) a
1033 cached ldap connection that the parent has called unbind on.
1035 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1036 of $address_data from the recipient verification was clobbered by the
1037 sender verification.
1039 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1040 was its contents. (It was OK if the option was not defined at all.)
1042 11. A "Completed" log line is now written for messages that are removed from
1043 the spool by the -Mrm option.
1045 12. $host_address is now set to the target address during the checking of
1046 ignore_target_hosts.
1048 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1049 being passed; this would have caused $sender_host_name to have been used if
1050 matching the list had actually called for a host name (not very likely,
1051 since this list is usually IP addresses). A host name is now passed as
1054 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1055 code that specifies a non-restarting handler (typically sigaction() in
1056 modern systems) in an attempt to fix a rare and obscure crash bug.
1058 15. Narrowed the window for a race in the daemon that could cause it to ignore
1059 SIGCHLD signals. This is not a major problem, because they are used only to
1060 wake it up if nothing else does.
1062 16. A malformed maildirsize file could cause Exim to calculate negative values
1063 for the mailbox size or file count. Odd effects could occur as a result.
1064 The maildirsize information is now recalculated if the size or filecount
1067 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1068 support for a long time. Removed HAVE_SYS_VFS_H.
1070 18. Updated exipick to current release from John Jetmore.
1072 19. Allow an empty sender to be matched against a lookup in an address list.
1073 Previously the only cases considered were a regular expression, or an
1076 20. Exim went into a mad DNS lookup loop when doing a callout where the
1077 host was specified on the transport, if the DNS lookup yielded more than
1080 21. The RFC2047 encoding function was originally intended for short strings
1081 such as real names; it was not keeping to the 75-character limit for
1082 encoded words that the RFC imposes. It now respects the limit, and
1083 generates multiple encoded words if necessary. To be on the safe side, I
1084 have increased the buffer size for the ${rfc2047: expansion operator from
1087 22. Failure to deliver a bounce message always caused it to be frozen, even if
1088 there was an errors_to setting on the router. The errors_to setting is now
1091 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1092 canonical form (fully expanded) before being placed in
1093 $sender_host_address.
1095 24. Updated eximstats to version 1.33
1097 25. Include certificate and key file names in error message when GnuTLS fails
1098 to set them up, because the GnuTLS error message doesn't include the name
1099 of the failing file when there is a problem reading it.
1101 26. Expand error message when OpenSSL has problems setting up cert/key files.
1104 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1105 (this can affect the format of dates).
1107 28. exim_tidydb, when checking for the continued existence of a message for
1108 which it has found a message-specific retry record, was not finding
1109 messages that were in split spool directories. Consequently, it was
1110 deleting retry records that should have stayed in existence.
1112 29. eximstats updated to version 1.35
1113 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1114 1.35 - bugfix such that pie charts by volume are generated correctly
1116 30. The SPA authentication driver was not abandoning authentication and moving
1117 on to the next authenticator when an expansion was forced to fail,
1118 contradicting the general specification for all authenticators. Instead it
1119 was generating a temporary error. It now behaves as specified.
1121 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1122 (the order specifies the preference for clients). The order is now AES256,
1123 AES128, 3DES, ARCFOUR128.
1125 31. Small patch to Sieve code - explicitly set From: when generating an
1128 32. Exim crashed if a remote delivery caused a very long error message to be
1129 recorded - for instance if somebody sent an entire SpamAssassin report back
1130 as a large number of 550 error lines. This bug was coincidentally fixed by
1131 increasing the size of one of Exim's internal buffers (big_buffer) that
1132 happened as part of the Exiscan merge. However, to be on the safe side, I
1133 have made the code more robust (and fixed the comments that describe what
1136 33. Some experimental protocols are using DNS PTR records for new purposes. The
1137 keys for these records are domain names, not reversed IP addresses. The
1138 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1139 leaves it alone. Component reversal etc. now happens only for IP addresses.
1142 34. The host_aton() function is supposed to be passed a string that is known
1143 to be a valid IP address. However, in the case of IPv6 addresses, it was
1144 not checking this. This is a hostage to fortune. Exim now panics and dies
1145 if the condition is not met. A case was found where this could be provoked
1146 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1147 components; fortuitously, this particular loophole had already been fixed
1148 by change 4.50/55 or 4.44/33 above.
1150 If there are any other similar loopholes, the new check in host_aton()
1151 itself should stop them being exploited. The report I received stated that
1152 data on the command line could provoke the exploit when Exim was running as
1153 exim, but did not say which command line option was involved. All I could
1154 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1155 running as the user.
1158 35. There was a buffer overflow vulnerability in the SPA authentication code
1159 (which came originally from the Samba project). I have added a test to the
1160 spa_base64_to_bits() function which I hope fixes it.
1163 36. The daemon start-up calls getloadavg() while still root for those OS that
1164 need the first call to be done as root, but it missed one case: when
1165 deliver_queue_load_max is set with deliver_drop_privilege. This is
1166 necessary for the benefit of the queue runner, because there is no re-exec
1167 when deliver_drop_privilege is set.
1169 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1170 was in use, was not putting the data itself into the right store pool;
1171 consequently, it could be overwritten for a subsequent message in the same
1172 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1175 38. Sometimes the final signoff response after QUIT could fail to get
1176 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1177 < 0 before doing a fflush(). This bug looks as though it goes back to the
1178 introduction of TLS in release 3.20, but "sometimes" must have been rare
1179 because the tests only now provoked it.
1185 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1186 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1187 Now it returns an integer. A number of calls were still expecting a T/F
1188 return. Fortuitously, in all cases, the tests worked in OK situations,
1189 which is the norm. However, things would have gone wrong on any write
1190 failures on the smtp file descriptor. This function is used when sending
1191 messages over SMTP and also when doing verify callouts.
1193 2. When Exim is called to do synchronous delivery of a locally submitted
1194 message (the -odf or -odi options), it no longer closes stderr before doing
1197 3. Implemented the mua_wrapper option.
1199 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1201 5. Implemented the functions header_remove(), header_testname(),
1202 header_add_at_position(), and receive_remove_recipient(), and exported them
1205 6. If an ACL "warn" statement specified the addition of headers, Exim already
1206 inserted X-ACL-Warn: at the start if there was no header name. However, it
1207 was not making this test for the second and subsequent header lines if
1208 there were newlines in the string. This meant that an invalid header could
1209 be inserted if Exim was badly configured.
1211 7. Allow an ACL "warn" statement to add header lines at the start or after all
1212 the Received: headers, as well as at the end.
1214 8. Added the rcpt_4xx retry error code.
1216 9. Added postmaster_mailfrom=xxx to callout verification option.
1218 10. Added mailfrom=xxxx to the callout verification option, for verify=
1221 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1222 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1224 12. Inserted some casts to stop certain compilers warning when using pointer
1225 differences as field lengths or precisions in printf-type calls (mostly
1226 affecting debugging statements).
1228 13. Added optional readline() support for -be (dynamically loaded).
1230 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1231 same clock tick as a message's arrival, so that its received time was the
1232 same as the "first fail" time on the retry record, and that message
1233 remained on the queue past the ultimate address timeout, every queue runner
1234 would try a delivery (because it was past the ultimate address timeout) but
1235 after another failure, the ultimate address timeout, which should have then
1236 bounced the address, did not kick in. This was a "< instead of <=" error;
1237 in most cases the first failure would have been in the next clock tick
1238 after the received time, and all would be well.
1240 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1241 being recognized when the domain list was tested by the match_domain
1242 condition in an expansion string.
1244 16. Added the ${str2b64: operator.
1246 17. Exim was always calling setrlimit() to set a large limit for the number of
1247 processes, without checking whether the existing limit was already
1248 adequate. (It did check for the limit on file descriptors.) Furthermore,
1249 errors from getrlimit() and setrlimit() were being ignored. Now they are
1250 logged to the main and panic logs, but Exim does carry on, to try to do its
1251 job under whatever limits there are.
1253 18. Imported PCRE 5.0.
1255 19. Trivial typo in log message " temporarily refused connection" (the leading
1258 20. If the log selector return_path_on_delivery was set and an address was
1259 redirected to /dev/null, the delivery process crashed because it assumed
1260 that a return path would always be set for a "successful" delivery. In this
1261 case, the whole delivery is bypassed as an optimization, and therefore no
1264 21. Internal re-arrangement: the function for sending a challenge and reading
1265 a response while authentication was assuming a zero-terminated challenge
1266 string. It's now changed to take a pointer and a length, to allow for
1267 binary data in such strings.
1269 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1271 23. Exim was not respecting finduser_retries when seeking the login of the
1272 uid under which it was called; it was always trying 10 times. (The default
1273 setting of finduser_retries is zero.) Also, it was sleeping after the final
1274 failure, which is pointless.
1276 24. Implemented tls_on_connect_ports.
1278 25. Implemented acl_smtp_predata.
1280 26. If the domain in control=submission is set empty, Exim assumes that the
1281 authenticated id is a complete email address when it generates From: or
1282 Sender: header lines.
1284 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1285 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1286 chown and chgrp in /bin and hostname in /usr/bin.
1288 28. Exim was keeping the "process log" file open after each use, just as it
1289 does for the main log. This opens the possibility of it remaining open for
1290 long periods when the USR1 signal hits a daemon. Occasional processlog
1291 errors were reported, that could have been caused by this. Anyway, it seems
1292 much more sensible not to leave this file open at all, so that is what now
1295 29. The long-running daemon process does not normally write to the log once it
1296 has entered its main loop, and it closes the log before doing so. This is
1297 so that log files can straightforwardly be renamed and moved. However,
1298 there are a couple of unusual error situations where the daemon does write
1299 log entries, and I had neglected to close the log afterwards.
1301 30. The text of an SMTP error response that was received during a remote
1302 delivery was being truncated at 512 bytes. This is too short for some of
1303 the long messages that one sometimes sees. I've increased the limit to
1306 31. It is now possible to make retry rules that apply only when a message has a
1307 specific sender, in particular, an empty sender.
1309 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1310 it possible to be selective about when SMTP synchronization is enforced.
1312 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1314 32. Implemented hosts_connection_nolog.
1316 33. Added an ACL for QUIT.
1318 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1321 35. Added mailbox_size and mailbox_filecount to appendfile.
1323 36. Added control = no_multiline_responses to ACLs.
1325 37. There was a bug in the logic of the code that waits for the clock to tick
1326 in the case where the clock went backwards by a substantial amount such
1327 that the microsecond fraction of "now" was more than the microsecond
1328 fraction of "then" (but the whole seconds number was less).
1330 38. Added support for the libradius Radius client library this is found on
1331 FreeBSD (previously only the radiusclient library was supported).
1337 1. When certain lookups returned multiple values in the form name=value, the
1338 quoting of the values was not always being done properly. Specifically:
1339 (a) If the value started with a double quote, but contained no whitespace,
1341 (b) If the value contained whitespace other than a space character (i.e.
1342 tabs or newlines or carriage returns) it was not quoted.
1343 This fix has been applied to the mysql and pgsql lookups by writing a
1344 separate quoting function and calling it from the lookup code. The fix
1345 should probably also be applied to nisplus, ibase and oracle lookups, but
1346 since I cannot test any of those, I have not disturbed their existing code.
1348 2. A hit in the callout cache for a specific address caused a log line with no
1349 reason for rejecting RCPT. Now it says "Previous (cached) callout
1350 verification failure".
1352 3. There was an off-by-one bug in the queryprogram router. An over-long
1353 return line was truncated at 256 instead of 255 characters, thereby
1354 overflowing its buffer with the terminating zero. As well as fixing this, I
1355 have increased the buffer size to 1024 (and made a note to document this).
1357 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1358 when Exim is waiting for an SMTP response from a remote server, Exim
1359 restarts its select() call on the socket, thereby resetting its timeout.
1360 This is not a problem when such interrupts are rare. Somebody set up a cron
1361 job to run exiwhat every 2 minutes, which is less than the normal select()
1362 timeout (5 or 10 minutes). This meant that the select() timeout never
1363 kicked in because it was always reset. I have fixed this by comparing the
1364 time when an interrupt arrives with the time at the start of the first call
1365 to select(). If more time than the timeout has elapsed, the interrupt is
1366 treated as a timeout.
1368 5. Some internal re-factoring in preparation for the addition of Sieve
1369 extensions (by MH). In particular, the "personal" test is moved to a
1370 separate function, and given an option for scanning Cc: and Bcc: (which is
1371 not set for Exim filters).
1373 6. When Exim created an email address using the login of the caller as the
1374 local part (e.g. when creating a From: or Sender: header line), it was not
1375 quoting the local part when it contained special characters such as @.
1377 7. Installed new OpenBSD configuration files.
1379 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1380 try to make them clearer.
1382 9. Callout options, other than the timeout value, were being ignored when
1383 verifying sender addresses in header lines. For example, when using
1385 verify = header_sender/callout=no_cache
1387 the cache was (incorrectly) being used.
1389 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1390 only the Cygwin environment.
1392 11. When return_path_on_delivery was set as a log selector, if different remote
1393 addresses in the same message used different return paths and parallel
1394 remote delivery occurred, the wrong values would sometimes be logged.
1395 (Whenever a remote delivery process finished, the return path value from
1396 the most recently started remote delivery process was logged.)
1398 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1399 lines when AUTH and/or TLS are in use. This is the "received protocol"
1400 field. Exim used to use "asmtp" for authenticated SMTP, without any
1401 indication (in the protocol name) for TLS use. Now it follows the RFC and
1402 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1403 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1404 names appear in log lines as well as in Received: header lines.
1406 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1407 extensions, and comparison tests, and to fix some bugs.
1409 14. Changes to the "personal" filter test:
1411 (1) The test was buggy in that it was just doing the equivalent of
1412 "contains" tests on header lines. For example, if a user's address was
1413 anne@some.where, the "personal" test would incorrectly be true for
1415 To: susanne@some.where
1417 This test is now done by extracting each address from the header in turn,
1418 and checking the entire address. Other tests that are part of "personal"
1419 are now done using regular expressions (for example, to check local parts
1420 of addresses in From: header lines).
1422 (2) The list of non-personal local parts in From: addresses has been
1423 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1424 taken from the Sieve specification recommendations.
1426 (3) If the message contains any header line starting with "List-" it is
1427 treated as non-personal.
1429 (4) The test for "circular" in the Subject: header line has been removed
1430 because it now seems ill-conceived.
1432 15. Minor typos in src/EDITME comments corrected.
1434 16. Installed latest exipick from John Jetmore.
1436 17. If headers_add on a router specified a text string that was too long for
1437 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1438 of string_sprintf() is now avoided.
1440 18. $message_body_size was not set (it was always zero) when running the DATA
1441 ACL and the local_scan() function.
1443 19. For the "mail" command in an Exim filter, no default was being set for
1444 the once_repeat time, causing a random time value to be used if "once" was
1445 specified. (If the value happened to be <= 0, no repeat happened.) The
1446 default is now 0s, meaning "never repeat". The "vacation" command was OK
1447 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1448 (I found it when inspecting the code).
1450 20. There is now an overall timeout for performing a callout verification. It
1451 defaults to 4 times the callout timeout, which applies to individual SMTP
1452 commands during the callout. The overall timeout applies when there is more
1453 than one host that can be tried. The timeout is checked before trying the
1454 next host. This prevents very long delays if there are a large number of
1455 hosts and all are timing out (e.g. when the network connections are timing
1456 out). The value of the overall timeout can be changed by specifying an
1457 additional sub-option for "callout", called "maxwait". For example:
1459 verify = sender/callout=5s,maxwait=20s
1461 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1462 to the end before writing, but this should make it even safer).
1464 22. Exim was forgetting that it had advertised PIPELINING for the second and
1465 subsequent messages on an SMTP connection. It was also not resetting its
1466 memory on STARTTLS and an internal HELO.
1468 23. When Exim logs an SMTP synchronization error within a session, it now
1469 records whether PIPELINING has been advertised or not.
1471 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1472 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1473 rather than long int.
1475 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1477 26. Added the never_mail option to autoreply.
1483 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1484 crash if the getsockname() call failed; this can happen if a connection is
1485 closed very soon after it is established. The problem was simply in the
1486 order in which certain operations were done, causing Exim to try to write
1487 to the SMTP stream before it had set up the file descriptor. The bug has
1488 been fixed by making things happen in the correct order.
1494 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1495 before the connection was closed, thus losing the rejection response.
1497 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1498 some early Solaris releases, but causes trouble in current releases where
1499 socklen_t is defined.
1501 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1504 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1507 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1508 that releasing the top of it at the end releases what was used for sub-
1509 expansions (unless the block got too big). However, discard this block if
1510 the first thing is a variable or header, so that we can use its block when
1511 it is dynamic (useful for very large $message_headers, for example).
1513 6. Lookups now cache *every* query, not just the most recent. A new, separate
1514 store pool is used for this. It can be recovered when all lookup caches are
1515 flushed. Lookups now release memory at the end of their result strings.
1516 This has involved some general refactoring of the lookup sources.
1518 7. Some code has been added to the store_xxx() functions to reduce the amount
1519 of flapping under certain conditions.
1521 8. log_incoming_interface used to affect only the <= reception log lines. Now
1522 it causes the local interface and port to be added to several more SMTP log
1523 lines, for example "SMTP connection from", and rejection lines.
1525 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1527 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1529 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1530 could be overwritten at the end of the current message (or the start of a
1531 new message if it was set in a HELO ACL). The value is now preserved for
1532 the duration of the SMTP connection.
1534 12. If a transport had a headers_rewrite setting, and a matching header line
1535 contained an unqualified address, that address was qualified, even if it
1536 did not match any rewriting rules. The underlying bug was that the values
1537 of the flags that permit the existence of unqualified sender and recipient
1538 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1539 non-local messages, and by -bnq for local messages) were not being
1540 preserved with the message after it was received.
1542 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1543 "next input=" as part of the text comprising the host identity instead of
1544 the correct text. The code was using the same buffer for two different
1545 strings. However, depending on which order the printing function evaluated
1546 its arguments, the bug did not always show up. Under Linux, for example, my
1547 test suite worked just fine.
1549 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1550 doesn't work with some older versions of Perl. It has been changed to "my",
1551 which in any case is probably the better facility to use.
1553 15. A really picky compiler found some instances of statements for creating
1554 error messages that either had too many or two few arguments for the format
1557 16. The size of the buffer for calls to the DNS resolver has been increased
1558 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1559 for addresses that have a lot of PTR records. This alleviates a problem; it
1560 does not fully solve it.
1562 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1563 buffer now truncates the list and logs the incident, which is the same
1564 action as happens when Exim is looking up a host name and its aliases.
1565 Previously in this situation something unpredictable would happen;
1566 sometimes it was "internal error: store_reset failed".
1568 18. If a server dropped the connection unexpectedly when an Exim client was
1569 using GnuTLS and trying to read a response, the client delivery process
1570 crashed while trying to generate an error log message.
1572 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1573 string, for example:
1575 warn message = H1: something\nH2: something
1577 the text was added as a single header line from Exim's point of view
1578 though it ended up OK in the delivered message. However, searching for the
1579 second and subsequent header lines using $h_h2: did not work. This has been
1580 fixed. Similarly, if a system filter added multiple headers in this way,
1581 the routers could not see them.
1583 20. Expanded the error message when iplsearch is called with an invalid key to
1584 suggest using net-iplsearch in a host list.
1586 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1587 ACLs are no longer actually imposed (and a message to that effect is
1590 22. If a "gecos" field in a passwd entry contained escaped characters, in
1591 particular, if it contained a \" sequence, Exim got it wrong when building
1592 a From: or a Sender: header from that name. A second bug also caused
1593 incorrect handling when an unquoted " was present following a character
1594 that needed quoting.
1596 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1597 was not being matched caselessly.
1599 24. Arranged for all hyphens in the exim.8 source to be escaped with
1602 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1603 properly. Recipient callout cache records were still being keyed to include
1604 the sender, even when use_sender was set false. This led to far more
1605 callouts that were necessary. The sender is no longer included in the key
1606 when use_sender is false.
1608 26. Added "control = submission" modifier to ACLs.
1610 27. Added the ${base62d: operator to decode base 62 numbers.
1612 28. dnsdb lookups can now access SRV records.
1614 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1615 the configuration file.
1617 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1618 (-v) mode. This makes the output for a verbose queue run more intelligible.
1620 31. Added a use_postmaster feature to recipient callouts.
1622 32. Added the $body_zerocount variable, containing the number of binary zero
1623 bytes in the message body.
1625 33. The time of last modification of the "new" subdirectory is now used as the
1626 "mailbox time last read" when there is a quota error for a maildir
1629 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1631 35. Added +ignore_unknown as a special item in host lists.
1633 36. Code for decoding IPv6 addresses in host lists is now included, even if
1634 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1635 address was recognized as an IP address, but was then not correctly decoded
1636 into binary, causing unexpected and incorrect effects when compared with
1643 1. Very minor rewording of debugging text in manualroute to say "list of
1644 hosts" instead of "hostlist".
1646 2. If verify=header_syntax was set, and a header line with an unqualified
1647 address (no domain) and a large number of spaces between the end of the
1648 name and the colon was received, the reception process suffered a buffer
1649 overflow, and (when I tested it) crashed. This was caused by some obsolete
1650 code that should have been removed. The fix is to remove it!
1652 3. When running in the test harness, delay a bit after writing a bounce
1653 message to get a bit more predictability in the log output.
1655 4. Added a call to search_tidyup() just before forking a reception process. In
1656 theory, someone could use a lookup in the expansion of smtp_accept_max_
1657 per_host which, without the tidyup, could leave open a database connection.
1659 5. Added the variables $recipient_data and $sender_data which get set from a
1660 lookup success in an ACL "recipients" or "senders" condition, or a router
1661 "senders" option, similar to $domain_data and $local_part_data.
1663 6. Moved the writing of debug_print from before to after the "senders" test
1666 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1667 problems for message scanning, either using a data ACL, or using
1668 local_scan() because the Received: header was not generated till after they
1669 were called (in order to set the time as the time of reception completion).
1670 I have revised the way this works. The header is now generated after the
1671 body is received, but before the ACL or local_scan() are called. After they
1672 are run, the timestamp in the header is updated.
1678 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1679 before starting a queue runner without re-exec. This happened only when
1680 deliver_drop_privilege was set or when the Exim user was set to root. The
1681 effect of the bug was that timeouts during subsequent deliveries caused
1682 crashes instead of being properly handled. The handler is now left at its
1683 default (and expected) setting.
1685 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1686 message, again when deliver_drop_privilege is set or Exim is run as root.
1687 The bug described in (1) was not present in this case, but the tidying up
1688 of the other signals was missing. I have made the two cases consistent.
1690 3. The ignore_target_hosts setting on a manualroute router was being ignored
1691 for hosts that were looked up using the /MX notation.
1693 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1696 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1697 operated on the sender address. After changing the $sender_address to <>
1698 for the sender address verify, Exim was re-instated it as the original
1699 (before rewriting) address, but remembering that it had rewritten it, so it
1700 wasn't rewriting it again. This bug also had the effect of breaking the
1701 sender address verification caching when the sender address was rewritten.
1703 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1704 This has been changed so that if the ip literal address matches
1705 ignore_target_hosts, the router declines.
1707 7. Added expansion conditions match_domain, match_address, and match_local_
1708 part (NOT match_host).
1710 8. The placeholder for the Received: header didn't have a length field set.
1712 9. Added code to Exim itself and to exim_lock to test for a specific race
1713 condition that could lead to file corruption when using MBX delivery. The
1714 issue is with the lockfile that is created in /tmp. If this file is removed
1715 after a process has opened it but before that process has acquired a lock,
1716 there is the potential for a second process to recreate the file and also
1717 acquire a lock. This could lead to two Exim processes writing to the file
1718 at the same time. The added code performs the same test as UW imapd; it
1719 checks after acquiring the lock that its file descriptor still refers to
1720 the same named file.
1722 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1723 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1726 11. Added the smtp_active_hostname option. If used, this will typically be made
1727 to depend on the incoming interface address. Because $interface_address is
1728 not set up until the daemon has forked a reception process, error responses
1729 that can happen earlier (such as "too many connections") no longer contain
1732 12. If an expansion in a condition on a "warn" statement fails because a lookup
1733 defers, the "warn" statement is abandoned, and the next ACL statement is
1734 processed. Previously this caused the whole ACL to be aborted.
1736 13. Added the iplsearch lookup type.
1738 14. Added ident_timeout as a log selector.
1740 15. Added tls_certificate_verified as a log selector.
1742 16. Added a global option tls_require_ciphers (compare the smtp transport
1743 option of the same name). This controls incoming TLS connections.
1745 17. I finally figured out how to make tls_require_ciphers do a similar thing
1746 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1747 before starting the TLS session.
1749 18. Tabs are now shown as \t in -bP output.
1751 19. If the log selector return_path_on_delivery was set, Exim crashed when
1752 bouncing a message because it had too many Received: header lines.
1754 20. If two routers both had headers_remove settings, and the first one included
1755 a superfluous trailing colon, the final name in the first list and the
1756 first name in the second list were incorrectly joined into one item (with a
1757 colon in the middle).
1763 1. Added -C and -D options to the exinext utility, mainly to make it easier
1764 to include in the automated testing, but these could be helpful when
1765 multiple configurations are in use.
1767 2. The exinext utility was not formatting the output nicely when there was
1768 an alternate port involved in the retry record key, nor when there was a
1769 message id as well (for retries that were specific to a specific message
1770 and a specific host). It was also confused by IPv6 addresses, because of
1771 the additional colons they contain. I have fixed the IPv4 problem, and
1772 patched it up to do a reasonable job for IPv6.
1774 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1775 delivery, the log line now contains "pipelined" if PIPELINING was used.
1777 4. An SMTP transport process used to panic and die if the bind() call to set
1778 an explicit outgoing interface failed. This has been changed; it is now
1779 treated in the same way as a connect() failure.
1781 5. A reference to $sender_host_name in the part of a conditional expansion
1782 that was being skipped was still causing a DNS lookup. This no longer
1785 6. The def: expansion condition was not recognizing references to header lines
1786 that used bh_ and bheader_.
1788 7. Added the _cache feature to named lists.
1790 8. The code for checking quota_filecount in the appendfile transport was
1791 allowing one more file than it should have been.
1793 9. For compatibility with Sendmail, the command line option
1801 and sets the incoming protocol and host name (for trusted callers). The
1802 host name and its colon can be omitted when only the protocol is to be set.
1803 Note the Exim already has two private options, -pd and -ps, that refer to
1804 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1805 "s", but I don't think that's a major issue.
1807 10. A number of refactoring changes to the code, none of which should affect
1810 (a) The number of logging options was getting close to filling up the
1811 32-bit word that was used as a bit map. I have split them into two classes:
1812 those that are passed in the argument to log_write(), and those that are
1813 only ever tested independently outside of that function. These are now in
1814 separate 32-bit words, so there is plenty of room for expansion again.
1815 There is no change in the user interface or the logging behaviour.
1817 (b) When building, for example, log lines, the code previously used a
1818 macro that called string_cat() twice, in order to add two strings. This is
1819 not really sufficiently general. Furthermore, there was one instance where
1820 it was actually wrong because one of the argument was used twice, and in
1821 one call a function was used. (As it happened, calling the function twice
1822 did not affect the overall behaviour.) The macro has been replaced by a
1823 function that can join an arbitrary number of extra strings onto a growing
1826 (c) The code for expansion conditions now uses a table and a binary chop
1827 instead of a serial search (which was left over from when there were very
1828 few conditions). Also, it now recognizes conditions like "pam" even when
1829 the relevant support is not compiled in: a suitably worded error message is
1830 given if an attempt is made to use such a condition.
1832 11. Added ${time_interval:xxxxx}.
1834 12. A bug was causing one of the ddress fields not to be passed back correctly
1835 from remote delivery subprocesses. The field in question was not being
1836 subsequently used, so this caused to problems in practice.
1838 13. Added new log selectors queue_time and deliver_time.
1840 14. Might have fixed a bug in maildirsizefile handling that threw up
1841 "unexpected character" debug warnings, and recalculated the data
1842 unnecessarily. In any case, I expanded the warning message to give more
1845 15. Added the message "Restricted characters in address" to the statements in
1846 the default ACL that block characters like @ and % in local parts.
1848 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1849 Three changes have been made:
1851 (a) There was a serious bug; a negative response to MAIL caused the whole
1852 recipient domain to be cached as invalid, thereby blocking all messages
1853 to all local parts at the same domain, from all senders. This bug has
1854 been fixed. The domain is no longer cached after a negative response to
1855 MAIL if the sender used is not empty.
1857 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1860 (c) A new callout option, "use_sender" has been added for people who want
1861 the modified behaviour.
1867 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1870 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1871 indeed breaks things for older releases.
1873 3. Added additional logging to the case where there is a problem reading data
1874 from a filter that is running in a subprocess using a pipe, in order to
1875 try to track down a specific problem.
1877 4. Testing facility fudge: when running in the test harness and attempting
1878 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1879 getting "No route to host". Convert this to a timeout.
1881 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1884 6. Some OS don't have socklen_t but use size_t instead. This affects the
1885 fifth argument of getsockopt() amongst other things. This is now
1886 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1887 can be set for individual OS. I have set it for SunOS5, OSF1, and
1888 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1889 some earlier ones do not.
1891 7. Change 4.30/15 was not doing the test caselessly.
1893 8. The standard form for an IPv6 address literal was being rejected by address
1894 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1895 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1896 this, as well as the form without the "IPv6" on the front (but only when
1897 address literals are enabled, of course).
1899 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1901 10. Exim crashed if a message with an empty sender address specified by -f
1902 encountered a router with an errors_to setting. This could be provoked only
1903 by a command such as
1907 where an empty string was supplied; "<>" did not hit this bug.
1909 11. Installed PCRE release 4.5.
1911 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1912 remained set. It is now erased.
1914 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1915 times from message ids (which are base 36 rather than the normal 62).
1917 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1918 were being counted as actual protocol errors, and logged if the log
1919 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1920 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1921 and DATA following a set of rejected RCPTs do not count as protocol errors.
1922 In other words, Exim assumes they were pipelined, though this may not
1923 actually be the case. Of course, in all cases the client gets an
1924 appropriate error code.
1926 15. If a lookup fails in an ACL condition, a message about the failure may
1927 be available; it is used if testing the ACL cannot continue, because most
1928 such messages specify what the cause of the deferral is. However, some
1929 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1930 that caused an old message to be retained and used if a later statement
1931 caused a defer, replacing the real cause of the deferral.
1933 16. If an IP address had so many PTR records that the DNS lookup buffer
1934 was not large enough to hold them, Exim could crash while trying to process
1935 the truncated data. It now detects and logs this case.
1937 17. Further to 4.21/58, another change has been made: if (and only if) the
1938 first line of a message (the first header line) ends with CRLF, a bare LF
1939 in a subsequent header line has a space inserted after it, so as not to
1940 terminate the header.
1942 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1943 unnecessarily, used atoi() instead of strtol(), and didn't check the
1944 termination when getting file sizes from file names by regex.
1946 19. Completely re-implemented the support for maildirsize files, in the light
1947 of a number of problems with the previous contributed implementation
1948 (4.30/29). In particular:
1950 . If the quota is zero, the maildirsize file is maintained, but no quota is
1953 . If the maildir directory does not exist, it is created before any attempt
1954 to write a maildirsize file.
1956 . The quota value in the file is just a cache; if the quota is changed in
1957 the transport, the new value overrides.
1959 . A regular expression is available for excluding directories from the
1962 20. The autoreply transport checks the characters in options that define the
1963 message's headers; it allows continued headers, but it was checking with
1964 isspace() after an embedded newline instead of explicitly looking for a
1967 21. If all the "regular" hosts to which an address was routed had passed their
1968 expiry times, and had not reached their retry times, the address was
1969 bounced, even if fallback hosts were defined. Now Exim should go on to try
1972 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1973 equivalent code in the SMTP transport. Some hosts send humungous responses
1974 to HELO/EHLO, more than 1024 it seems.
1976 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1977 gives compiler warnings in some environments. I've now done it "properly",
1980 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1981 (because of problems with the built-in one) was declared to return uschar *
1982 instead of char *, causing compiler failure.
1984 25. Fixed a file descriptor leak when processing alias/forward files.
1986 26. Fixed a minor format string issue in dbfn.c.
1988 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1990 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1991 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1994 29. When a sender address is verified, it is cached, to save repeating the test
1995 when there is more than one recipient in a message. However, when the
1996 verification involves a callout, it is possible for different callout
1997 options to be set for different recipients. It is too complicated to keep
1998 track of this in the cache, so now Exim always runs a verification when a
1999 callout is required, relying on the callout cache for the optimization.
2000 The overhead is duplication of the address routing, but this should not be
2003 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2004 to be verified with callout=postmaster, and the main callout worked but the
2005 postmaster check failed, the verification correctly failed. However, if a
2006 subsequent RCPT command asked for sender verification *without* the
2007 postmaster check, incorrect caching caused this verification also to fail,
2010 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2011 it was not caching the DNS options (qualify_single, search_parents) that
2012 were used when the lookup failed. A subsequent lookup with different
2013 options therefore always gave the same answer, though there were cases
2014 where it should not have. (Example: a "domains = !$mx_any" option on a
2015 dnslookup router: the "domains" option is always processed without any
2016 widening, but the router might have qualify_single set.) Now Exim uses the
2017 cached value only when the same options are set.
2019 32. Added John Jetmore's "exipick" utility to the distribution.
2021 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2022 than a timeout (e.g. a certificate is required, and is not provided), an
2023 Exim server now closes the connection immediately. Previously it waited for
2024 the client to close - but if the client is SSL, it seems that they each
2025 wait for each other, leading to a delay before one of them times out.
2027 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2028 maintained 0.8.x compatibility because I don't think many are using it, and
2029 it is clearly obsolete.
2031 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2034 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2035 client certificate was expired. A simple patch fixes this, though I don't
2036 understand the full logic of why the verify callback is called multiple
2039 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2040 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2041 which causes problems with some clients (such as the Certicom SSL Plus
2042 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2043 disables the coutermeasure allowing Eudora to connect."
2045 38. Exim was not checking that a write() to a log file succeeded. This could
2046 lead to Bad Things if a log got too big, in particular if it hit a file
2047 size limit. Exim now panics and dies if it cannot write to a log file, just
2048 as it does if it cannot open a log file.
2050 39. Modified OS/Makefile-Linux so that it now contains
2052 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2054 The two -D definitions ensure that Exim is compiled with large file
2055 support, which makes it possible to handle log files that are bigger than
2058 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2059 instance) a domain was checked against a named list that involved a lookup,
2060 causing $domain_data to be set, then another domain was checked against the
2061 same list, then the first domain was re-checked, the value of $domain_data
2062 after the final check could be wrong. In particular, if the second check
2063 failed, it could be set empty. This bug probably also applied to
2066 41. The strip_trailing_dot option was not being applied to the address given
2067 with the -f command-line option.
2069 42. The code for reading a message's header from the spool was incrementing
2070 $received_count, but never initializing it. This meant that the value was
2071 incorrect (doubled) while delivering a message in the same process in which
2072 it was received. In the most common configuration of Exim, this never
2073 happens - a fresh exec is done - but it can happen when
2074 deliver_drop_privilege is set.
2076 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2077 it now includes up to 150 characters of the unexpected data in the log
2080 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2081 and building data strings. The size of both of these buffers was 10 000
2082 bytes - far larger than anybody would *ever* want, thought I. Needless to
2083 say, somebody hit the limit. I have increased the maximum line length to
2084 20 000 and the maximum data length of concatenated lines to 100 000. I have
2085 also fixed two bugs, because there was no checking on these buffers. Tsk,
2086 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2087 buffer is too small.
2089 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2090 lsearch lookups. Now it does.
2092 46. When parsing a route_list item in a manualroute router, a fixed-length
2093 buffer was used for the list of hosts. I made this 1024 bytes long,
2094 thinking that nobody would ever have a list of hosts that long. Wrong.
2095 Somebody had a whole pile of complicated expansion conditions, and the
2096 string was silently truncated, leading to an expansion error. It turns out
2097 that it is easier to change to an unlimited length (owing to other changes
2098 that have happened since this code was originally written) than to build
2099 structure for giving a limitation error. The length of the item that
2100 expands into the list of hosts is now unlimited.
2102 47. The lsearch lookup could not handle data where the length of text line was
2103 more than 4095 characters. Such lines were truncated, leading to shortened
2104 data being returned. It should now handle lines of any length.
2106 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2107 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2110 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2111 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2112 no longer happens in comment lines. A list of replacements is now placed
2113 at the head of all of the source files, except those whose only change is
2114 to replace PERL_COMMAND in the very first #! line.
2116 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2117 messages on the queue, with a bottom-up merge sort, using code contributed
2118 by Michael Haardt. This should make operations like -bp somewhat faster on
2119 large queues. It won't affect queue runners, except when queue_run_in_order
2122 51. Installed eximstats 1.31 in the distribution.
2124 52. Added support for SRV lookups to the dnslookup router.
2126 53. If an ACL referred to $message_body or $message_body_end, the value was not
2127 reset for any messages that followed in the same SMTP session.
2129 54. The store-handling optimization for building very long strings was not
2130 differentiating between the different store pools. I don't think this
2131 actually made any difference in practice, but I've tidied it.
2133 55. While running the routers to verify a sender address, $sender_address
2134 was still set to the sender address. This is wrong, because when routing to
2135 send a bounce to the sender, it would be empty. Therefore, I have changed
2136 it so that, while verifying a sender address, $sender_address is set to <>.
2137 (There is no change to what happens when verifying a recipient address.)
2139 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2140 target A or AAAA records (if not already returned) without resetting the
2141 qualify_single or search_parents options of the DNS resolver. These are
2142 inappropriate in this case because the targets of MX and SRV records must
2143 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2144 target that, when qualified, matched something in the local domain. These
2145 two options are now turned off when doing these lookups.
2147 57. It seems that at least some releases of Reiserfs (which does not have the
2148 concept of a fixed number of inodes) returns zero and not -1 for the
2149 number of available inodes. This interacted badly with check_spool_inodes,
2150 which assumed that -1 was the "no such thing" setting. What I have done is
2151 to check that the total number of inodes is greater than zero before doing
2152 the test of how many are available.
2154 58. When a "warn" ACL statement has a log_message modifier, the message is
2155 remembered, and not repeated. This is to avoid a lot of repetition when a
2156 message has many recipients that cause the same warning to be written.
2157 Howewer, Exim was preserving the list of already written lines for an
2158 entire SMTP session, which doesn't seem right. The memory is now reset if a
2159 new message is started.
2161 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2162 debugging output unless log_rewrite was also set.
2164 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2165 of (int)(handle) when we know that handle contains (void *)(-1).
2167 61. The Exim daemon panic-logs an error return when it closes the incoming
2168 connection. However "connection reset by peer" seems to be common, and
2169 isn't really an error worthy of noting specially, so that particular error
2172 62. When Exim is trying to find all the local interfaces, it used to panic and
2173 die if the ioctl to get the interface flags failed. However, it seems that
2174 on at least one OS (Solaris 9) it is possible to have an interface that is
2175 included in the list of interfaces, but for which you get a failure error
2176 for this call. This happens when the interface is not "plumbed" into a
2177 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2178 failure of the "get flags" call assumes that the interface is down.
2180 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2181 makes life easier for people who are doing arithmetic on fields extracted
2182 from dates, where you often get leading zeros that should not be
2183 interpreted as octal.
2185 64. Added qualify_domain to the redirect router, to override the global
2188 65. If a pathologically long header line contained very many addresses (the
2189 report of this problem mentioned 10 000) and each of them was rewritten,
2190 Exim could use up a very large amount of memory. (It kept on making new
2191 copies of the header line as it rewrote, and never released the old ones.)
2192 At the expense of a bit more processing, the header rewriting function has
2193 been changed so that it no longer eats memory in this way.
2195 66. The generation of the Received: header has been moved from the time that a
2196 message starts to be received, to the time that it finishes. The timestamp
2197 in the Received: header should now be very close to that of the <= log
2198 line. There are two side-effects of this change:
2200 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2201 logged header lines no longer include the local Received: line, because
2202 it has not yet been created. The same applies to a copy of the message
2203 that is returned to a non-SMTP sender when a message is rejected.
2205 (b) When a filter file is tested using -bf, no additional Received: header
2206 is added to the test message. After some thought, I decided that this
2209 This change does not affect the value of $received_for. It is still set
2210 after address rewriting, but before local_scan() is called.
2212 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2214 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2215 gave an unhelpful panic error message, and a defer error. I have managed to
2216 change this behaviour so that it now rejects any supplied certificate,
2217 which seems right, as the list of acceptable certificates is empty.
2219 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2220 gave an unhelpful defer error. I have not managed to make this reject any
2221 supplied certificates, but the error message it gives is "no certificate
2222 supplied", which is not helpful.
2224 70. exigrep's output now also includes lines that are not associated with any
2225 message, but which match the given pattern. Implemented by a patch from
2226 Martin Sluka, which also tidied up the Perl a bit.
2228 71. Recipient callout verification, like sender verification, was using <> in
2229 the MAIL FROM command. This isn't really the right thing, since the actual
2230 sender may affect whether the remote host accepts the recipient or not. I
2231 have changed it to use the actual sender in the callout; this means that
2232 the cache record is now keyed on a recipient/sender pair, not just the
2233 recipient address. There doesn't seem to be a real danger of callout loops,
2234 since a callout by the remote host to check the sender would use <>.
2235 [SEE ABOVE: changed after hitting problems.]
2237 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2238 temporary errors. However, in the case of such a code being given after
2239 the end of a data transmission (i.e. after ".") Exim was failing to write
2240 a retry record for the message. (Yes, there was some broken host that was
2241 actually sending 8xx at this point.)
2243 73. An unknown lookup type in a host list could cause Exim to panic-die when
2244 the list was checked. (An example that provoked this was putting <; in the
2245 middle of a list instead of at the start.) If this happened during a DATA
2246 ACL check, a -D file could be left lying around. This kind of configuration
2247 error no longer causes Exim to die; instead it causes a defer errror. The
2248 incident is still logged to the main and panic logs.
2250 74. Buglet left over from Exim 3 conversion. The message "too many messages
2251 in one connection" was written to the rejectlog but not the mainlog, except
2252 when address rewriting (yes!) was being logged.
2254 75. Added write_rejectlog option.
2256 76. When a system filter was run not as root (that is, when system_filter_user
2257 was set), the values of the $n variables were not being returned to the
2258 main process; thus, they were not subsequently available in the $sn
2261 77. Added +return_path_on_delivery log selector.
2263 78. A connection timeout was being treated differently from recipients deferred
2264 when testing hosts_max_try with a message that was older than the host's
2265 retry timeout. (The host should not be counted, thus allowing all hosts to
2266 be tried at least once before bouncing.) This may have been the cause of an
2267 occasionally reported bug whereby a message would remain on the queue
2268 longer than the retry timeout, but would be bounced if a delivery was
2269 forced. I say "may" because I never totally pinned down the problem;
2270 setting up timeout/retry tests is difficult. See also the next item.
2272 79. The ultimate address timeout was not being applied to errors that involved
2273 a combination of host plus message (for example, a timeout on a MAIL
2274 command). When an address resolved to a number of possible hosts, and they
2275 were not all tried for each delivery (e.g. because of hosts_max_try), a
2276 message could remain on the queue longer than the retry timeout.
2278 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2281 81. Fixed an obscure SMTP outgoing bug which required at least the following
2282 conditions: (a) there was another message waiting for the same server;
2283 (b) the server returned 5xx to all RCPT commands in the first message so
2284 that the message was not completed; (c) the server dropped the connection
2285 or gave a negative response to the RSET that Exim sends to abort the
2286 transaction. The observed case was a dropped connection after DATA that had
2287 been sent in pipelining mode. That is, the server had advertised PIPELINING
2288 but was not implementing it correctly. The effect of the bug was incorrect
2289 behaviour, such as trying another host, and this could lead to a crash.
2295 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2296 and daemon.c were passed as pointers to ints; they should have been
2297 pointers to socklen_t variables (which are typically unsigned ints).
2299 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2302 3. Fixed a really odd bug that affected only the testing scheme; patching a
2303 certain fixed string in the binary changed the value of another string that
2304 happened to be identical to the end of the original first string.
2306 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2307 name", it returns that address as the IP address. On some operating
2308 systems (e.g. Solaris), it also passes back the IP address string as the
2309 "host name". However, on others (e.g. Linux), it passes back an empty
2310 string. Exim wasn't checking for this, and was changing the host name to an
2311 empty string, assuming it had been canonicized.
2313 5. Although rare, it is permitted to have more than one PTR record for a given
2314 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2315 all the names associated with an address, because they do in Solaris.
2316 However, it seems that they do not in Linux for data that comes from the
2317 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2318 I found this out when I moved to a new Linux workstation and tried to run
2319 the Exim test suite.
2321 To get round this problem I have changed the code so that it now does its
2322 own call to the DNS to look up PTR records when searching for a host name.
2323 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2324 addresses that are only in /etc/hosts are still found.
2326 This behaviour is, however, controlled by an option called host_lookup_
2327 order, which defaults to "bydns:byaddr". If people want to use the other
2328 order, or indeed, just use one or the other means of lookup, they can
2329 specify it in this variable.
2331 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2332 some operating systems, this comes back from gethostbyaddr() as an empty
2333 string, and this is what Exim used to test for. However, it seems that in
2334 other systems, "." is yielded. Exim now tests for this case too.
2336 7. The values of check_spool_space and check_log_space are now held internally
2337 as a number of kilobytes instead of an absolute number of bytes. If a
2338 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2339 kilobyte. This means that much larger values can be stored.
2341 8. Exim monitor: an attempt to get the action menu when not actually pointing
2342 at a message produces an empty menu entitled "No message selected". This
2343 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2344 no entries in it ("Shell widget menu has zero width and/or height"). So I
2345 have added a single, blank menu entry in this case.
2347 9. Added ${quote_local_part.
2349 10. MIME decoding is now applied to the contents of Subject: header lines when
2352 11. Now that a reference to $sender_host_address automatically causes a reverse
2353 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2354 host lookup before query-style lookups in lists that might use this
2355 variable. This has therefore been abolished, and the "net-" prefix is no
2356 longer necessary for query-style lookups.
2358 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2359 have been a typo for LFLAGS, so it has been changed.
2361 13. The install script calls Exim with "-C /dev/null" in order to find the
2362 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2363 to be output. Howeve, since Exim outputs its version number before the
2364 error, it didn't break the script. It just looked ugly. I fixed this by
2365 always allowing "-C /dev/null" if the caller is root.
2367 14. Ignore overlarge ACL variable number when reading spool file - insurance
2368 against a later release with more variables having written the file.
2370 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2371 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2372 this, as well as the form without the "IPv6" on the front.
2374 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2375 OS/Makefile-Darwin file.
2377 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2378 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2380 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2381 variable. However, some OS use ints for inodes. I've added cast to long int
2382 to get rid of the compiler warning.
2384 19. I had forgotten to lock out "/../" in configuration file names when
2385 ALT_CONFIG_PREFIX was set.
2387 20. Routers used for verification do not need to specify transports. However,
2388 if such a router generated a host list, and callout was configured, Exim
2389 crashed, because it could not find a port number from the (non-existent)
2390 transport. It now assumes port 25 in this circumstance.
2392 21. Added the -t option to exigrep.
2394 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2395 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2396 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2397 them is set, LOOKUP_LSEARCH is forced.
2399 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2401 24. Added sender and host information to the "rejected by local_scan()" log
2402 line; previously there was no indication of these.
2404 25. Added .include_if_exists.
2406 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2407 Linux. It turns out that not all file systems support this. Apparently some
2408 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2409 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2410 not supported on the file descriptor, is now ignored when Exim is trying to
2411 sync a directory. This applies only to Linux.
2413 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2415 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2416 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2419 29. Added support for maildirsize files from supplied patch (modified a bit).
2421 30. The use of :fail: followed by an empty string could lead Exim to respond to
2422 sender verification failures with (e.g.):
2424 550 Verification failed for <xxx>
2425 550 Sender verify failed
2427 where the first response line was missing the '-' that indicates it is not
2428 the final line of the response.
2430 31. The loop for finding the name of the user that called Exim had a hardwired
2431 limit of 10; it now uses the value of finduser_retries, which is used for
2432 all other user lookups.
2434 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2437 33. Exim was neglecting to zero errno before one call of strtol() when
2438 expanding a string and expecting an integer value. On some systems this
2439 resulted in spurious "integer overflow" errors. Also, it was casting the
2440 result into an int without checking.
2442 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2443 did not work. The code looks as if it has *never* worked, though it appears
2444 to have been documented since at least releast 1.62. I have made it work.
2446 35. The "timeout_DNS" error in retry rules, also documented since at least
2447 1.62, also never worked. As it isn't clear exactly what this means, and
2448 clearly it isn't a major issue, I have abolished the feature by treating it
2449 as "timeout", and writing a warning to the main and panic logs.
2451 36. The display of retry rules for -brt wasn't always showing the error code
2454 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2455 timeout_connect_A, timeout_connect_MX.
2457 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2458 to the empty sender.
2460 39. The daemon was not analysing the content of -oX till after it had closed
2461 stderr and disconnected from the controlling terminal. This meant that any
2462 syntax errors were only noted on the panic log, and the return code from
2463 the command was 0. By re-arranging the code a little, I've made the
2464 decoding happen first, so such errors now appear on stderr, and the return
2465 code is 1. However, the actual setting up of the sockets still happens in
2466 the disconnected process, so errors there are still only recorded on the
2469 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2470 connections (as happens on some IP stacks) was logged at start up time as
2471 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2472 it from "IPv6 and IPv4", which means that two separate sockets are being
2475 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2476 says whether AF_INET or AF_INET6 was passed as an argument.
2478 42. Exiwhat output was messed up when time zones were included in log
2481 43. Exiwhat now gives more information about the daemon's listening ports,
2482 and whether -tls-on-connect was used.
2484 44. The "port" option of the smtp transport is now expanded.
2486 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2487 silently ignored. Now an error message is written to the main and panic
2490 46. There's a new ACL modifier called "logwrite" which writes to a log file
2491 as soon as it is encountered.
2493 47. Added $local_user_uid and $local_user_gid at routing time.
2495 48. Exim crashed when trying to verify a sender address that was being
2498 49. Exim was recognizing only a space character after ".include". It now also
2499 recognizes a tab character.
2501 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2502 extracting the relevant information from the specification. The man page no
2503 longer contains scrambled data for the -d option, and I've added a section
2504 at the front about calling Exim under different names.
2506 51. Added "extra_headers" argument to the "mail" command in filter files.
2508 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2511 53. Installed eximstats 1.29.
2513 54. Added transport_filter_timeout as a generic transport option.
2515 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2516 Cc: header lines. This was required by RFC 822, but it not required by RFC
2519 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2520 incoming messages that did not have them. Now it does so only if the
2521 message originates locally, that is, if there is no associated remote host
2522 address. When Resent- header lines are present, this applies to the Resent-
2523 lines rather than the non-Resent- lines.
2525 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2526 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2528 58. Messages for configuration errors now include the name of the main
2529 configuration file - useful now that there may be more than one file in a
2530 list (.included file names were always shown).
2532 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2533 for those rare installations that do not start the daemon as root or run it
2534 setuid root. I've cut out the call to initgroups() if the daemon is not
2537 60. The Exim user and group can now be bound into the binary as text strings
2538 that are looked up at the start of Exim's processing.
2540 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2542 62. Added $mailstore_basename variable.
2544 63. Installed patch to sieve.c from Michael Haardt.
2546 64. When Exim failed to open the panic log after failing to open the main log,
2547 the original message it was trying to log was written to stderr and debug
2548 output, but if they were not available (the usual case in production), it
2549 was lost. Now it is written to syslog before the two lines that record the
2550 failures to open the logs.
2552 65. Users' Exim filters run in subprocesses under the user's uid. It is
2553 possible for a "deliver" command or an alias in a "personal" command to
2554 provoke an address rewrite. If logging of address rewriting is configured,
2555 this fails because the process is not running as root or exim. There may be
2556 a better way of dealing with this, but for the moment (because 4.30 needs
2557 to be released), I have disabled address rewrite logging when running a
2558 filter in a non-root, non-exim process.
2564 1. The buildconfig auxiliary program wasn't quoting the value set for
2565 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2566 not defined. This bug was masked in 4.22 by the effect that was fixed in
2569 2. Some messages that were rejected after a message id was allocated were
2570 shown as "incomplete" by exigrep. It no longer does this for messages that
2571 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2573 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2574 have allow_domain_literals set, the ID did not get logged in the <= line.
2575 Domain literals are now always recognized in Message-ID: header lines.
2577 4. The first argument for a ${extract expansion item is the key name or field
2578 number. Leading and trailing spaces in this item were not being ignored,
2579 causing some misleading effects.
2581 5. When deliver_drop_privilege was set, single queue runner processes started
2582 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2583 same command in the process it spins off) were not dropping privilege.
2585 6. When the daemon running as "exim" started a queue runner, it always
2586 re-executed Exim in the spun-off process. This is a waste of effort when
2587 deliver_drop_privilege is set. The new process now just calls the
2588 queue-runner function directly.
2594 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2597 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2598 ignored. Though the use of -oP was forcing the writing of a pid file, it
2599 was always written to the default place.
2601 3. If the message "no IP address found for host xxxx" is generated during
2602 incoming verification, it is now followed by identification of the incoming
2603 connection (so you can more easily find what provoked it).
2605 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2607 5. Added some features to "harden" Exim a bit more against certain attacks:
2609 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2610 be put in Local/Makefile. This is like the never_users runtime option,
2611 but it cannot be overridden. The default setting is "root".
2613 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2614 prefix string with which any file named in a -C command line option
2617 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2618 is retained for -C and -D only if the caller of Exim is root. Without
2619 it, the exim user may also use -C and -D and retain privilege.
2621 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2622 command line option is disabled.
2624 6. Macro names set by the -D option must start with an upper case letter, just
2625 like macro names defined in the configuration file.
2627 7. Added "dereference=" facility to LDAP.
2629 8. Two instances of the typo "uknown" in the source files are fixed.
2631 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2632 the Configure-Makefile script screwed up while processing it.
2634 10. Incorporated PCRE 4.4.
2636 11. The SMTP synchronization check was not operating right at the start of an
2637 SMTP session. For example, it could not catch a HELO sent before the client
2638 waited for the greeting. There is now a check for outstanding input at the
2639 point when the greeting is written. Because of the duplex, asynchronous
2640 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2641 way, but not yet received, when the check is performed.
2643 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2644 on TCP/IP sockets, because this apparently causes some broken clients to
2647 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2648 unchanged) from the Cygwin maintainer.
2650 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2651 was supported instead of testing for mbx. Effectively a typo.
2653 15. The spa authenticator server code was not checking that the input it
2654 received was valid base64.
2656 16. The debug output line for the "set" modifier in ACLs was not showing the
2657 name of the variable that was being set.
2659 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2661 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2662 lookup on its own. Something else was needed to trigger the lookup. For
2663 example, a match in host_lookup or the need for a host name in a host list.
2664 Now, if $sender_host_name is referenced and the host name has not yet been
2665 looked up, a lookup is performed. If the lookup fails, the variable remains
2666 empty, and $host_lookup_failed is set to "1".
2668 19. Added "eqi" as a case-independent comparison operator.
2670 20. The saslauthd authentication condition could segfault if neither service
2671 nor realm was specified.
2673 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2674 error message that was logged was misleading, and incoming SMTP
2675 connections were dropped. The message is now more accurate, and temporary
2676 errors are given to SMTP connections.
2678 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2679 (except RSET) with a 421 error, until QUIT is received. However, it was
2680 failing to send a response to QUIT.
2682 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2683 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2684 ACL is now run after the helo_try_verify_hosts code.
2686 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2687 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2688 case-independent, so other case variants are also recognized). Apparently
2689 some systems use these upper case variants.
2691 25. If more than two messages were waiting for the same host, and a transport
2692 filter was specified for the transport, Exim sent two messages over the
2693 same TCP/IP connection, and then failed with "socket operation on non-
2694 socket" when it tried to send the third.
2696 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2698 27. The extern definition of crypt16() in expand.c was not being excluded when
2699 the OS had its own crypt16() function.
2701 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2702 as a preferred synonym for return_size_limit, both as an option and as an
2705 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2707 30. Changed the default configuration ACL to relax the local part checking rule
2708 for addresses that are not in any local domains. For these addresses,
2709 slashes and pipe symbols are allowed within local parts, but the sequence
2710 /../ is explicitly forbidden.
2712 31. SPA server authentication was not clearing the challenge buffer before
2715 32. log_message in a "warn" ACL statement was writing to the reject log as
2716 well as to the main log, which contradicts the documentation and doesn't
2717 seem right (because no rejection is happening). So I have stopped it.
2719 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2720 However, I am unable to do any testing of this.
2722 34. Fixed an infelicity in the appendfile transport. When checking directories
2723 for a mailbox, to see if any needed to be created, it was accidentally
2724 using path names with one or more superfluous leading slashes; tracing
2725 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2727 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2728 contents are added to the log line that is written for every discarded
2729 recipient. (Previously a log_message setting was ignored.)
2731 36. The ${quote: operator now quotes the string if it is empty.
2733 37. The install script runs exim in order to find its version number. If for
2734 some reason other than non-existence or emptiness, which it checks, it
2735 could not run './exim', it was installing it with an empty version number,
2736 i.e. as "exim-". This error state is now caught, and the installation is
2739 38. An argument was missing from the function that creates an error message
2740 when Exim fails to connect to the socket for saslauthd authentication.
2741 This could cause Exim to crash, or give a corrupted message.
2743 39. Added isip, isip4, and isip6 to ${if conditions.
2745 40. The ACL variables $acl_xx are now saved with the message, and can be
2746 accessed later in routers, transports, and filters.
2748 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2749 strings in the file are not string-expanded.
2751 42. If a MAIL command specified a SIZE value that was too large to fit into an
2752 int variable, the check against message_size_limit failed. Such values are
2753 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2754 day this will have to be increased, but I don't think I want to be around
2755 when emails are that large.
2762 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2763 iconv() is not standard in FreeBSD.
2765 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2766 IPv6 enabled. The observed symptom was a segmentation fault on return from
2767 the function os_common_find_running_interfaces() in src/os.c.
2769 3. In the check_special_case() function in daemon.c I had used "errno" as an
2770 argument name, which causes warnings on some systems. This was basically a
2771 typo, since it was named "eno" in the comments!
2773 4. The code that waits for the clock to tick (at a resolution of some fraction
2774 of a second) so as to ensure message-id uniqueness was always waiting for
2775 at least one whole tick, when it could have waited for less. [This is
2776 almost certainly not relevant at current processor speeds, where it is
2777 unlikely to ever wait at all. But we try to future-proof.]
2779 5. The function that sleeps for a time interval that includes fractions of a
2780 second contained a race. It did not block SIGALRM between setting the
2781 timer, and suspending (a couple of lines later). If the interval was short
2782 and the sigsuspend() was delayed until after it had expired, the suspension
2783 never ended. On busy systems this could lead to processes getting stuck for
2786 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2787 process, before it forks any delivery processes. The open lookup caching
2788 mechanism meant that the open file or database connection was passed into
2789 the delivery process. The problem was that delivery processes always tidy
2790 up cached lookup data. This could cause a problem for the next delivery
2791 process started by the queue runner, because the external queue runner
2792 process does not know about the closure. So the next delivery process
2793 still has data in the lookup cache. In the case of a file lookup, there was
2794 no problem because closing a file descriptor in a subprocess doesn't affect
2795 the parent. However, if the lookup was caching a connection to a database,
2796 the connection was closed, and the second delivery process was likely to
2797 see errors such as "PGSQL: query failed: server closed the connection
2798 unexpectedly". The problem has been fixed by closing all cached lookups
2799 in a queue runner before running a delivery process.
2801 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2802 seem to have the "const" qualifier which it has on other OS. I've
2805 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2806 *of the same type* (client or server) with the same public name that an
2807 error should be diagnosed.
2809 9. When Exim looked up a host name for an IP address, but failed to find the
2810 original IP address when looking up the host name (a safety check), it
2811 output the message "<ip address> does not match any IP for NULL", which was
2812 confusing, to say the least. The bug was that the host name should have
2813 appeared instead of "NULL".
2815 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2816 user that is built into the binary, and the -C or -D options is used, root
2817 privilege is dropped before the configuration file is read. In addition,
2818 logging is switched to stderr instead of the normal log files. If the
2819 configuration then re-defines the Exim user, the unprivileged environment
2820 is probably not what is expected, so Exim logs a panic warning message (but
2823 However, if deliver_drop_privilege is set, the unprivileged state may well
2824 be exactly what is intended, so the warning has been cut out in that case,
2825 and Exim is allowed to try to write to its normal log files.
2831 1. smtp_return_error_details was not giving details for temporary sender
2832 or receiver verification errors.
2834 2. Diagnose a configuration error if two authenticators have the same public
2837 3. Exim used not to create the message log file for a message until the first
2838 delivery attempt. This could be confusing when incoming messages were held
2839 for policy or load reasons. The message log file is now created at the time
2840 the message is received, and an initial "Received" line is written to it.
2842 4. The automatically generated man page for command line options had a minor
2843 bug that caused no ill effects; however, a more serious problem was that
2844 the procedure for building the man page automatically didn't always
2845 operate. Consequently, release 4.20 contains an out-of-date version. This
2846 shouldn't happen again.
2848 5. When building Exim with embedded Perl support, the script that builds the
2849 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2850 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2852 6. The freeze_tell option was not being used for messages that were frozen on
2853 arrival, either by an ACL or by local_scan().
2855 7. Added the smtp_incomplete_transaction log selector.
2857 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2858 was accepting AUTH without a new EHLO.
2860 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2861 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2864 10. Logging of TCP/IP connections (when configured) now happens in the main
2865 daemon process instead of the child process, so that the TCP/IP connection
2866 count is more accurate (but it can never be perfect).
2868 11. The use of "drop" in a nested ACL was not being handled correctly in the
2869 outer ACL. Now, if condition failure induced by the nested "drop" causes
2870 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2871 or "require"), the connection is dropped.
2873 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2874 that yield "discard" can now be used with an "accept" or a "discard" verb,
2875 but an error is generated for any others (because I can't see a useful way
2876 to define what should happen).
2878 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2879 are now processed in the same way as lines in the Exim configuration file.
2880 In particular, continuation lines are supported.
2882 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2884 15. Added -ti meaning -t -i.
2886 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2887 domains, and warn by logging if others are found.
2889 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2890 returned from GIFCONF when Exim is trying to find the list of interfaces on
2891 a host. The code in os.c has been modified to copy each ifreq to an aligned
2892 structure in all cases.
2894 Also, in some cases, the returned ifreq's were being copied to a 'struct
2895 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2896 means the last couple of bytes of an IPv6 address could be chopped if the
2897 ifreq contained only a normal sockaddr (14 bytes storage).
2899 18. Named domain lists were not supported in the hosts_treat_as_local option.
2900 An entry such as +xxxx was not recognized, and was treated as a literal
2903 19. Ensure that header lines added by a DATA ACL are included in the reject log
2904 if the ACL subsequently rejects the message.
2906 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2907 MD5 (which is deprecated).
2909 21. When testing a filter file using -bf, Exim was writing a message when it
2910 took the sender from a "From " line in the message, but it was not doing so
2911 when it took $return_path from a Return-Path: header line. It now does.
2913 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2914 with a valid header line field name (a series of printing characters
2915 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2917 23. Changed "disc" in the source to "disk" to conform to the documentation and
2918 the book and for uniformity.
2920 24. Ignore Sendmail's -Ooption=value command line item.
2922 25. When execve() failed while trying to run a command in a pipe transport,
2923 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2924 could be confused with a return value of 69 from the command itself. This
2925 has been changed to 127, the value the shell returns if it is asked to run
2926 a non-existent command. The wording for the related log line suggests a
2927 non-existent command as the problem.
2929 26. If received_header_text expands to an empty string, do not add a Received:
2930 header line to the message. (Well, it adds a token one on the spool, but
2931 marks it "old" so that it doesn't get used or transmitted.)
2933 27. Installed eximstats 1.28 (addition of -nt option).
2935 28. There was no check for failure on the call to getsockname() in the daemon
2936 code. This can fail if there is a shortage of resources on the system, with
2937 ENOMEM, for example. A temporary error is now given on failure.
2939 29. Contrary to the C standard, it seems that in some environments, the
2940 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2941 program. Exim now does this explicitly; it affects the formatting of
2942 timestamps using strftime().
2944 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2945 complaints. I've now initialized all the variables, to avoid this.
2947 32. Header lines added by a system filter were not being "seen" during
2948 transport-time rewrites.
2950 33. The info_callback() function passed to OpenSSL is set up with type void
2951 (*)(SSL *, int, int), as described somewhere. However, when calling the
2952 function (actually a macro) that sets it up, the type void(*)() is
2953 expected. I've put in a cast to prevent warnings from picky compilers.
2955 34. If a DNS black list lookup found a CNAME record, but there were no A
2956 records associated with the domain it pointed at, Exim crashed.
2958 35. If a DNS black list lookup returned more than one A record, Exim ignored
2959 all but the first. It now scans all returned addresses if a particular IP
2960 value is being sought. In this situation, the contents of the
2961 $dnslist_value variable are a list of all the addresses, separated by a
2964 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2965 to accept a host name and all its aliases if the forward lookup for any of
2966 them yielded the IP address of the incoming connection. Now it accepts only
2967 those names whose forward lookup yields the correct IP address. Any other
2968 names are discarded. This closes a loophole whereby a rogue DNS
2969 administrator could create reverse DNS records to break through a
2970 wildcarded host restriction in an ACL.
2972 37. If a user filter or a system filter that ran in a subprocess used any of
2973 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2974 the wrong values were passed to the pipe command ($thisaddress had the
2975 value of $0, $0 had the value of $1, etc). This bug was introduced by
2976 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2978 38. Improved the line breaking for long SMTP error messages from ACLs.
2979 Previously, if there was no break point between 40 and 75 characters, Exim
2980 left the rest of the message alone. Two changes have been made: (a) I've
2981 reduced the minimum length to 35 characters; (b) if it can't find a break
2982 point between 35 and 75 characters, it looks ahead and uses the first one
2983 that it finds. This may give the occasional overlong line, but at least the
2984 remaining text gets split now.
2986 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2987 file descriptors might be low, and that setting 1000 would always raise it.
2988 It turns out that in some environments, the limit is already over 1000 and
2989 that lowering it causes trouble. So now Exim takes care not to decrease it.
2991 40. When delivering a message, the value of $return_path is set to $sender_
2992 address at the start of routing (routers may change the value). By an
2993 oversight, this default was not being set up when an address was tested by
2994 -bt or -bv, which affected the outcome if any router or filter referred to
2997 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2998 log only when "message" or "log_message" are set. However, if one of the
2999 conditions was an address verification, or a call to a nested ACL, the
3000 messages generated by the underlying test were being passed through. This
3001 no longer happens. The underlying message is available in $acl_verify_
3002 message for both "message" and "log_message" expansions, so it can be
3003 passed through if needed.
3005 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3006 new expansion $bh_ to give the encoded byte string without charset
3007 translation. Translation happens only if iconv() is available; HAVE_ICONV
3008 indicates this at build time. HEADERS_CHARSET gives the charset to
3009 translate to; headers_charset can change it in the configuration, and
3010 "headers charset" can change it in an individual filter file.
3012 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3013 that creates RFC 2047 encoded "words" labels them as that charset instead
3014 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3015 expansion operator; (ii) when Exim creates a From: line for a local
3016 message; (iii) when a header line is rewritten to include a "phrase" part.
3018 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3019 buggy, causing it to skip the first lines of messages whose message ID
3020 ended in 'D'. This would not have bitten before Exim release 4.14, because
3021 message IDs were unlikely to end in 'D' before then. The effect was to have
3022 incorrect size information for certain domains.
3024 45. #include "config.h" was missing at the start of the crypt16.c module. This
3025 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3028 46. If there was a timeout during a "random" callout check, Exim treated it as
3029 a failure of the random address, and carried on sending RSET and the real
3030 address. If the delay was just some slowness somewhere, the response to the
3031 original RCPT would be taken as a response to RSET and so on, causing
3032 mayhem of various kinds.
3034 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3035 when I implemented it. It didn't allow for the fact that some option values
3036 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3037 the right test for positive values.
3039 48. Domain names in DNS records are case-independent. Exim always looks them up
3040 in lower case. Some resolvers return domain names in exactly the case they
3041 appear in the zone file, that is, they may contain uppercase letters. Not
3042 all resolvers do this - some return always lower case. Exim was treating a
3043 change of case by a resolver as a change of domain, similar to a widening
3044 of a domain abbreviation. This triggered its re-routing code and so it was
3045 trying to route what was effectively the same domain again. This normally
3046 caused routing to fail (because the router wouldn't handle the domain
3047 twice). Now Exim checks for this case specially, and just changes the
3048 casing of the domain that it ultimately uses when it transmits the message
3051 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3054 50. If a filter generated a file delivery with a non-absolute name (possible if
3055 no home directory exists for the router), the forbid_file option was not
3058 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3059 the existing equality matching.
3061 52. Exim was using ints instead of ino_t variables in some places where it was
3062 dealing with inode numbers.
3064 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3065 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3066 called TMPDIR, and if it finds it is different, it changes its value.
3068 54. The smtp_printf() function is now made available to local_scan() so
3069 additional output lines can be written before returning. There is also an
3070 smtp_fflush() function to enable the detection of a dropped connection.
3071 The variables smtp_input and smtp_batched_input are exported to
3074 55. Changed the default runtime configuration: the message "Unknown user"
3075 has been removed from the ACL, and instead placed on the localuser router,
3076 using the cannot_route_message feature. This means that any verification
3077 failures that generate their own messages won't get overridden. Similarly,
3078 the "Unrouteable address" message that was in the ACL for unverifiable
3079 relay addresses has also been removed.
3081 56. Added hosts_avoid_esmtp to the smtp transport.
3083 57. The exicyclog script was not checking for the esoteric option
3084 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3085 will work only if exicyclog is run under the appropriate euid.
3087 58. Following a discussion on the list, the rules by which Exim recognises line
3088 endings on incoming messages have been changed. The -dropcr and drop_cr
3089 options are now no-ops, retained only for backwards compatibility. The
3090 following line terminators are recognized: LF CRLF CR. However, special
3091 processing applies to CR:
3093 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3094 nor a local message in the state where . is a terminator.
3096 (ii) If a bare CR is encountered in a header line, an extra space is added
3097 after the line terminator so as not to end the header. The reasoning
3098 behind this is that bare CRs in header lines are most likely either
3099 to be mistakes, or people trying to play silly games.
3101 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3102 was being incorrectly given as 18 bytes larger than it should have been.
3103 This is a VOB (very old bug).
3105 60. This may never have affected anything current, but just in case it has:
3106 When the local host is found other than at the start of a list of hosts,
3107 the local host, those with the same MX, and any that follow, are discarded.
3108 When the list in question was part of a longer list of hosts, the following
3109 hosts (not currently being processed) were also being discarded. This no
3110 longer happens. I'm not sure if this situation could ever has previously
3113 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3116 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3117 header. This is something that is recommended in a new Internet Draft, and
3118 is something that is documented as being done by Sendmail. There are two
3119 possible values. For messages generated by the autoreply transport, Exim
3122 Auto-Submitted: auto-replied
3124 whereas for all other generated messages (e.g. bounces) it adds
3126 Auto-Submitted: auto-generated
3128 63. The "personal" condition in filters now includes a test for the
3129 Auto-Submitted: header. If it contains the string "auto-" the message it
3130 not considered personal.
3132 64. Added rcpt_include_affixes as a generic transport option.
3134 65. Added queue_only_override (default true).
3136 66. Added the syslog_duplication option.
3138 67. If what should have been the first header line of a message consisted of
3139 a space followed by a colon, Exim was mis-interpreting it as a header line.
3140 It isn't of course - it is syntactically invalid and should therefore be
3141 treated as the start of the message body. The misbehaviour could have
3142 caused a number of strange effects, including loss of data in subsequent
3143 header lines, and spool format errors.
3145 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3146 client host had authenticated. This control can now be exercised by an ACL
3147 for more flexibility.
3149 69. By default, callouts do not happen when testing with -bh. There is now a
3150 variant, -bhc, which does actually run the callout code, including
3151 consulting and updating the callout cache.
3153 70. Added support for saslauthd authentication, courtesy of Alexander
3156 71. If statvfs() failed on the spool or log directories while checking their
3157 size for availability, Exim confusingly gave the error "space shortage".
3158 Furthermore, in debugging mode it crashed with a floating point exception.
3159 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3160 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3161 serious problem, Exim now writes to the main and panic logs when this
3162 happens, with details of the failure. It then refuses to accept the
3163 incoming message, giving the message "spool directory problem" or "log
3164 directory problem" with a 421 code for SMTP messages.
3166 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3167 0, 1, and 2 exist, because some OS complain for execs without them (see
3168 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3169 descriptors. However, the code omitted to check that the open succeeded,
3170 causing mysterious errors if for some reason the permissions on /dev/null
3171 got screwed. Now Exim writes a message to the main and panic logs, and
3172 bombs out if it can't open /dev/null.
3174 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3175 interact so that it is all more flexible. It is supposed to remain
3176 backwards compatible. Also added extra_local_interfaces.
3178 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3179 to bomb out with an assertion failure - to the client this appears as a
3180 connection drop. This problem occurs in the part of the code that was taken
3181 from the Samba project. Fortunately, the assertion is in a very simple
3182 function, so I have fixed this by reproducing the function inline in the
3183 one place where it is called, and arranging for authentication to fail
3184 instead of killing the process with assert().
3186 75. The SPA client code was not working when the server requested OEM rather
3187 than Unicode encoding.
3189 76. Added code to make require_files with a specific uid setting more usable in
3190 the case where statting the file as root fails - usually a non-root-mounted
3191 NFS file system. When this happens and the failure is EACCES, Exim now
3192 forks a subprocess and does the per-uid checking as the relevant uid.
3194 77. Added process_log_path.
3196 78. If log_file_path was not explicitly set, a setting of check_log_space or
3197 check_log_inodes was ignored.
3199 79. If a space check for the spool or log partitions fails, the incident is now
3200 logged. Of course, in the latter case the data may get lost...
3202 80. Added the %p formatting code to string_format() so that it can be used to
3203 print addresses in debug_print(). Adjusted all the address printing in the
3204 debugging in store.c to use %p rather than %d.
3206 81. There was a concern that a line of code in smtp_in.c could overflow a
3207 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3208 initially expressed, the concern was not well-founded, because trailing
3209 spaces are removed early. However, if the trailing spaces were followed by
3210 a NULL, they did not get removed, so the overflow was possible. Two fixes
3213 (a) I re-wrote the offending code in a cleaner fashion.
3214 (b) If an incoming SMTP command contains a NULL character, it is rejected
3217 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3218 runs initgroups(), so that if the Exim user is in any additional groups,
3219 they will be used during message reception.
3225 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob