1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.213 2005/08/30 10:07:58 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Added the "success_on_redirect" address verification option. See
10 NewStuff for rationale and an example.
12 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
14 PH/02 Patch to exigrep to allow it to work on syslog lines.
16 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
17 fread() to skip over the body file's header line, because in Cygwin the
18 header line is locked and is inaccessible.
20 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
21 co-exist for some time) to make it clear that it is the Exim ID that is
22 referenced, not the Message-ID: header line.
24 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
25 string_format() function, because snprintf() does not exist on all
28 PH/06 The use of forbid_filter_existstest now also locks out the use of the
29 ${stat: expansion item.
31 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
32 protocol synchronization error", to keep the pedants happy.
34 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
35 well as for IRIX systems, when gcc is being used. See the host.c source
38 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
40 PH/10 Named domain lists were not working if used in a queue_smtp_domains
43 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
44 transport and to the smtp transport in LMTP mode.
46 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
48 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
49 run a filter in a subprocess. This could lead to confusion in subsequent
50 lookups in the parent process. There should also be a search_tidyup() at
51 the end of the subprocess.
53 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
54 only if the host matched helo_try_verify_hosts, which caused the
55 verification to occur when the EHLO/HELO command was issued. The ACL just
56 tested the remembered result. Now, if a previous verification attempt has
57 not happened, "verify = helo" does it there and then.
59 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
61 TK/03 Fix log output including CR from clamd.
63 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
64 exist provoked a memory error which could cause a segfault.
66 PH/15 Installed PCRE 6.2
68 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
70 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
71 of the problem. Specifically, suggested +O2 rather than +O1 for the
74 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
76 PH/20 If a delivery was routed to a non-standard port by means of an SRV
77 record, the port was not correctly logged when the outgoing_port log
78 selector was set (it logged the transort's default port).
80 PH/21 Added support for host-specific ports to manualroute, queryprogram,
81 fallback_hosts, and "hosts" in the smtp transport.
83 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
84 host errors such as "Connection refused".
86 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
87 authentication with radiusclient 0.4.9:
89 - Error returned from rc_read_config was caught wrongly
90 - Username/password not passed on to radius server due to wrong length.
92 The presumption is that some radiusclient API changes for 4.51/PH/17
93 were not taken care of correctly. The code is still untested by me (my
94 Linux distribution still has 0.3.2 of radiusclient), but it was
95 contributed by a Radius user.
97 PH/24 When doing a callout, the value of $domain wasn't set correctly when
98 expanding the "port" option of the smtp transport.
100 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
101 while reading a MIME header. Thanks to Tom Hughes for a patch.
103 PH/24 Include config.h inside local_scan.h so that configuration settings are
106 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
107 that in an ACL for RCPT (for example), you can examine exactly what was
110 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
111 commands, but it was not correctly comparing the address with the actual
112 client host address. Thus, it would show the EHLO address in Received:
113 header lines when this was not necessary.
115 PH/27 Added the % operator to ${eval:}.
117 PH/28 Exim tries to create and chdir to its spool directory when it starts;
118 it should be ignoring failures (because with -C, for example, it has lost
119 privilege). It wasn't ignoring creation failures other than "already
122 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
125 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
126 because an input line was too long, either on its own, or by virtue of
127 too many continuations, the temporary file was not being removed.
129 PH/31 Missing "BOOL" in function definition in filtertest.c.
135 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
137 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
138 command itself ended in error, the underlying message about the transport
139 filter timeout was being overwritten with the pipe command error. Now the
140 underlying error message should be appended to the second error message.
142 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
144 PH/02 Reduce the amount of output that "make" produces by default. Full output
145 can still be requested.
147 PH/03 The warning log line about a condition test deferring for a "warn" verb
148 was being output only once per connection, rather than after each
149 occurrence (because it was using the same function as for successful
150 "warn" verbs). This seems wrong, so I have changed it.
152 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
153 it should not have, which might have caused a crash in the right
154 circumstances, but probably never did.
156 PH/04 Installed a modified version of Tony Finch's patch to make submission
157 mode fix the return path as well as the Sender: header line, and to
158 add a /name= option so that you can make the user's friendly name appear
161 TF/03 Added the control = fakedefer ACL modifier.
163 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
164 Mark Lowes for thorough testing.
166 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
168 TK/03 Merged latest SRS patch from Miles Wilton.
170 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
171 with the definition in sysexits.h (which is #included earlier).
172 Fortunately, Exim does not actually use EX_OK. The code used to try to
173 preserve the sysexits.h value, by assumimg that macro definitions were
174 scanned for macro replacements. I have been disabused of this notion,
175 so now the code just undefines EX_OK before #including unistd.h.
177 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
178 in the smtp transport. When a block could not be written in a single
179 write() function, the timeout was being re-applied to each part-write.
180 This seems wrong - if the receiver was accepting one byte at a time it
181 would take for ever. The timeout is now adjusted when this happens. It
182 doesn't have to be particularly precise.
184 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
185 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
187 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
188 without a domain if the check to <postmaster@domain> fails.
190 SC/01 Eximstats: added -xls and the ability to specify output files
191 (patch written by Frank Heydlauf).
193 SC/02 Eximstats: use FileHandles for outputing results.
195 SC/03 Eximstats: allow any combination of xls, txt, and html output.
197 SC/04 Eximstats: fixed display of large numbers with -nvr option
199 SC/05 Eximstats: fixed merging of reports with empty tables.
201 SC/06 Eximstats: added the -include_original_destination flag
203 SC/07 Eximstats: removed tabs and trailing whitespace.
205 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
207 TK/06 MBOX spool code: Add real "From " MBOX separator line
208 so the .eml file is really in mbox format (even though
209 most programs do not really care). Patch from Alex Miller.
211 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
212 The latter is generated from $received_to and is only set if the
213 message has one envelope recipient. SA can use these headers,
214 obviously out-of-the-box. Patch from Alex Miller.
216 PH/08 The ${def test on a variable was returning false if the variable's
217 value was "0", contrary to what the specification has always said!
218 The result should be true unless the variable is empty.
220 PH/09 The syntax error of a character other than { following "${if
221 def:variable_name" (after optional whitespace) was not being diagnosed.
222 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
223 accidental colon was present, for example, could give incorrect results.
225 PH/10 Tidied the code in a number of places where the st_size field of a stat()
226 result is used (not including appendfile, where other changes are about
229 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
230 This involved changing a lot of size variables from int to off_t. It
231 should work with maildirs and everything.
233 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
234 spamd dying while we are connected to it.
236 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
239 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
240 function for BDB 4.3.
242 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
244 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
245 the sources for PCRE have been re-arranged and more files are now
248 PH/15 The code I had for printing potentially long long variables in PH/11
249 above was not the best (it lost precision). The length of off_t variables
250 is now inspected at build time, and an appropriate printing format (%ld
251 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
252 to be "long long int" or "long int". This is needed for the internal
253 formatting function string_vformat().
255 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
256 the configuration file to be ":syslog", then the script "guesses" where
257 the logs files are, rather than using the compiled in default. In our
258 case the guess is not the same as the compiled default, so the script
259 suddenly stopped working when I started to use syslog. The patch checks
260 to see if log_file_path is "". If so, it attempts to read it from exim
261 with no configuration file to get the compiled in version, before it
262 falls back to the previous guessing code."
264 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
265 implementing BATV in an Exim configuration. See NewStuff for the gory
268 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
269 Makefile that are specific to HP-UX.
271 PH/18 If the "use_postmaster" option was set for a recipient callout together
272 with the "random" option, the postmaster address was used as the MAIL
273 FROM address for the random test, but not for the subsequent recipient
274 test. It is now used for both.
276 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
277 patch removes a few documentation additions to RFC 3028, because the
278 latest draft now contains them. It adds the new en;ascii-case comparator
279 and a new error check for 8bit text in MIME parts. Comparator and
280 require names are now matched exactly. I enabled the subaddress
281 extension, but it is not well tested yet (read: it works for me)."
283 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
284 rework some of the code of TK/09 above to avoid the hardwired use of
285 "%lld" and "long long". Replaced the call to snprintf() with a call to
288 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
289 records point to non-existent hosts", "retry timeout exceeded", and
290 "retry time not reached for any host after a long failure period".
292 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
293 experimental DomainKeys support:
295 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
296 (2) On an error such as an illegally used "control", the wrong name for
297 the control was given.
299 These problems did NOT occur unless DomainKeys support was compiled.
301 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
303 PH/24 Added ${if match_ip condition.
305 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
306 databases so that it will be absolutely obvious if a crash occurs in the
307 DB library. This is a regular occurrence (often caused by mis-matched
310 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
311 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
312 picked up on a user's system that detects such things. There doesn't seem
313 to be a gcc warning option for this - only an attribute that has to be
314 put on the function's prototype. It seems that in Fedora Core 4 they have
315 set this on a number of new functions. No doubt there will be more in due
318 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
319 specify a transport. However, if an address that was verified by such a
320 router was the subject of a callout, Exim crashed because it tried to
321 read the rcpt_include_affixes from the non-existent transport. Now it
322 just assumes that the setting of that option is false. This bug was
323 introduced by 4.51/PH/31.
325 PH/28 Changed -d+all to exclude +memory, because that information is very
326 rarely of interest, but it makes the output a lot bigger. People tend to
327 do -d+all out of habit.
329 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
330 code in os-type was giving problems when libc.so lives in lib64, like on
333 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
334 aren't the modern standard, and it seems that some systems' include files
335 don't always have them. Exim was already checking for some of the newer
336 ones like T_AAAA, and defining it itself. I've added checks for all the
337 record types that Exim uses.
339 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
340 not automatically generating a new one, as it is supposed to. This
341 prevented TLS from working. If the file did exist, but contained invalid
342 data, a new version was generated, as expected. It was only the case of a
343 non-existent file that was broken.
345 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
346 with a change in libdomainkeys > 0.64.
348 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
349 from DNS. If the selector record carries the flag, it now has
350 precedence over the domain-wide flag.
352 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
354 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
355 the use of a port name that isn't defined in /etc/services) occurred, the
356 message was deferred as in a normal delivery, and thus remained on the
357 spool, instead of being failed because of the mua_wrapper setting. This
358 is now fixed, and I tidied up some of the mua_wrapper messages at the
361 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
362 the messages in a hash of arrays rather than using individual hashes.
363 This is a bit cleaner and results in dramatic memory savings, albeit
364 at a slight CPU cost.
366 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
367 as requested by Marc Sherman.
369 SC/10 Eximstats: added histograms for user specified patterns as requested
372 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
374 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
375 fopen() in the content-scanning modules that did not already have it.
381 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
382 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
384 TK/02 Fix ACL "control" statement not being available in MIME ACL.
386 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
388 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
389 to test Sieve filters that use "vacation".
391 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
392 that changes the way the GnuTLS parameters are stored in the cache file.
393 The new format can be generated externally. For backward compatibility,
394 if the data in the cache doesn't make sense, Exim assumes it has read an
395 old-format file, and it generates new data and writes a new file. This
396 means that you can't go back to an older release without removing the
399 PH/03 A redirect router that has both "unseen" and "one_time" set does not
400 work if there are any delivery delays because "one_time" forces the
401 parent to be marked "delivered", so its unseen clone is never tried
402 again. For this reason, Exim now forbids the simultaneous setting of
405 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
406 redirected to themselves ("homonym" addresses). Read the long ChangeLog
407 entry if you want to know the details. The fix, however, neglected to
408 consider the case when local delivery batching is involved. The test for
409 "previously delivered" was not happening when checking to see if an
410 address could be batched with a previous (undelivered) one; under
411 certain circumstances this could lead to multiple deliveries to the same
414 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
415 in its include files, and this causes problems building Exim.
417 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
418 header_syntax) but Exim was just ignoring anything given after a slash.
419 In particular, this caused confusion with an attempt to use "verify =
420 reverse_host_lookup/defer_ok". An error is now given when options are
421 supplied for verify items that do not have them. (Maybe reverse_host_
422 lookup should have a defer_ok option, but that's a different point.)
424 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
425 defined by RFC 821) to 2048, because there were problems with some AUTH
426 commands, and RFC 1869 says the size should be increased for extended
427 SMTP commands that take arguments.
429 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
432 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
433 "unknown" error; now it says that the functionality isn't in the binary.
435 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
436 an address' error message when a string expansion fails (syntax or
437 whatever). Otherwise the password may appear in the log. Following change
438 PH/42 below, there is no longer a chance of it appearing in a bounce
441 PH/11 Installed exipick version 20050225.0 from John Jetmore.
443 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
444 of its addresses was ever tried. (Bugzilla bug #2.)
446 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
447 the result incorrectly in the debug output. (It correctly added a newline
448 to what was transported.)
450 TF/01 Added $received_time.
452 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
453 commented out examples of how to interface to a virus scanner and to
454 SpamAssassin. Also added commented examples of av_scanner and
455 spamd_address settings.
457 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
458 and controls are allowed in which ACLs. There were a couple of minor
459 errors. Some of the entries in the conditions table (which is a table of
460 where they are NOT allowed) were getting very unwieldy; rewrote them as a
461 negation of where the condition IS allowed.
463 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
465 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
466 header file does not have a version number, so I've had to invent a new
467 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
468 API. The code is untested by me (my Linux distribution still has 0.3.2 of
469 radiusclient), but it was contributed by a Radius user.
471 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
472 files or directories, for OpenSSL.
474 PH/19 When an Exim process that is running as root has to create an Exim log
475 file, it does so in a subprocess that runs as exim:exim so as to get the
476 ownership right at creation (otherwise, other Exim processes might see
477 the file with the wrong ownership). There was no test for failure of this
478 fork() call, which would lead to the process getting stuck as it waited
479 for a non-existent subprocess. Forks do occasionally fail when resources
480 run out. I reviewed all the other calls to fork(); they all seem to check
483 PH/20 When checking for unexpected SMTP input at connect time (before writing
484 the banner), Exim was not dealing correctly with a non-positive return
485 from the read() function. If the client had disconnected by this time,
486 the result was a log entry for a synchronization error with an empty
487 string after "input=" when read() returned zero. If read() returned -1
488 (an event I could not check), uninitialized data bytes were printed.
489 There were reports of junk text (parts of files, etc) appearing after
492 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
494 PH/22 Added support for macro redefinition, and (re)definition in between
495 driver and ACL definitions.
497 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
498 forgetting to use the resulting value; it was using the unexpanded value.
500 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
501 hadn't been configured. The fix is from Juergen Kreileder, who
502 understands it better than I do:
504 "Here's what I see happening with three configured cyrus_sasl
505 authenticators configured (plain, login, cram-md5):
507 On startup auth_cyrus_sasl_init() gets called for each of these.
508 This means three calls to sasl_listmech() without a specified mech_list.
509 => SASL tests which mechs of all available mechs actually work
510 => three warnings about OTP not working
511 => the returned list contains: plain, login, cram-md5, digest-md5, ...
513 With the patch, sasl_listmech() also gets called three times. But now
514 SASL's mech_list option is set to the server_mech specified in the the
515 authenticator. Or in other words, the answer from sasl_listmech()
516 gets limited to just the mech you're testing for (which is different
518 => the return list contains just 'plain' or 'login', 'cram-md5' or
519 nothing depending on the value of ob->server_mech.
521 I've just tested the patch: Authentication still works fine,
522 unavailable mechs specified in the exim configuration are still
523 caught, and the auth.log warnings about OTP are gone."
525 PH/25 When debugging is enabled, the contents of the command line are added
526 to the debugging output, even when log_selector=+arguments is not
529 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
530 answer is "GNU", and only if the return is "GNU/something" is the answer
533 PH/27 $acl_verify_message is now set immediately after the failure of a
534 verification in an ACL, and so is available in subsequent modifiers. In
535 particular, the message can be preserved by coding like this:
537 warn !verify = sender
538 set acl_m0 = $acl_verify_message
540 Previously, $acl_verify_message was set only while expanding "message"
541 and "log_message" when a very denied access.
543 PH/28 Modified OS/os.c-Linux with
545 -#ifndef OS_LOAD_AVERAGE
546 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
548 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
549 nomenclature these days.)
551 PH/29 Installed patch from the Sieve maintainer that adds the options
552 sieve_useraddress and sieve_subaddress to the redirect router.
554 PH/30 In these circumstances:
555 . Two addresses routed to the same list of hosts;
556 . First host does not offer TLS;
557 . First host accepts first address;
558 . First host gives temporary error to second address;
559 . Second host offers TLS and a TLS session is established;
560 . Second host accepts second address.
561 Exim incorrectly logged both deliveries with the TLS parameters (cipher
562 and peerdn, if requested) that were in fact used only for the second
565 PH/31 When doing a callout as part of verifying an address, Exim was not paying
566 attention to any local part prefix or suffix that was matched by the
567 router that accepted the address. It now behaves in the same way as it
568 does for delivery: the affixes are removed from the local part unless
569 rcpt_include_affixes is set on the transport.
571 PH/32 Add the sender address, as F=<...>, to the log line when logging a
572 timeout during the DATA phase of an incoming message.
574 PH/33 Sieve envelope tests were broken for match types other than :is. I have
575 applied a patch sanctioned by the Sieve maintainer.
577 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
578 the uid or gid is negative. A case of a negative gid caused this to be
579 noticed. The fix allows for either to be negative.
581 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
582 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
585 PH/36 The change PH/12 above was broken. Fixed it.
587 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
588 the grounds that routing the same address twice would always produce the
589 same answer. This might have been true once, but it is certainly no
590 longer true now. Routing a child address may depend on the previous
591 routing that produced that child. Some complicated redirection strategies
592 went wrong when messages had multiple recipients, and made Exim's
593 behaviour dependent on the order in which the addresses were given.
595 I have moved the duplicate checking until after the routing is complete.
596 Exim scans the addresses that are assigned to local and remote
597 transports, and removes any duplicates. This means that more work will be
598 done, as duplicates will always all be routed, but duplicates are
599 presumably rare, so I don't expect this is of any significance.
601 For deliveries to pipes, files, and autoreplies, the duplicate checking
602 still happens during the routing process, since they are not going to be
605 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
606 It corrects a timeout issue with spamd. This is Ian's comment: "The
607 background is that sometimes spamd either never reads data from a
608 connection it has accepted, or it never writes response data. The exiscan
609 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
610 blindly assumes that writes won't block so it may never time out."
612 PH/39 Allow G after quota size as well as K and M.
614 PH/40 The value set for $authenticated_id in an authenticator may not contain
615 binary zeroes or newlines because the value is written to log lines and
616 to spool files. There was no check on this. Now the value is run through
617 the string_printing() function so that such characters are converted to
618 printable escape sequences.
620 PH/41 $message_linecount is a new variable that contains the total number of
621 lines in the message. Compare $body_linecount, which is the count for the
624 PH/42 Exim no longer gives details of delivery errors for specific addresses in
625 bounce and delay warning messages, except in certain special cases, which
628 (a) An SMTP error message from a remote host;
629 (b) A message specified in a :fail: redirection;
630 (c) A message specified in a "fail" command in a system filter;
631 (d) A message specified in a FAIL return from the queryprogram router;
632 (e) A message specified by the cannot_route_message router option.
634 In these cases only, Exim does include the error details in bounce and
635 warning messages. There are also a few cases where bland messages such
636 as "unrouteable address" or "local delivery error" are given.
638 PH/43 $value is now also set for the "else" part of a ${run expansion.
640 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
641 being worked on, but at least Exim now implements the latest version to
644 PH/45 In a pipe transport, although a timeout while waiting for the pipe
645 process to complete was treated as a delivery failure, a timeout while
646 writing the message to the pipe was logged, but erroneously treated as a
647 successful delivery. Such timeouts include transport filter timeouts. For
648 consistency with the overall process timeout, these timeouts are now
649 treated as errors, giving rise to delivery failures by default. However,
650 there is now a new Boolean option for the pipe transport called
651 timeout_defer, which, if set TRUE, converts the failures into defers for
652 both kinds of timeout. A transport filter timeout is now identified in
655 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
656 systems where "make" and "gmake" are different, calling "gmake" at top
657 level broke things. I've arranged for the value of $(MAKE) to be passed
658 from the Makefile to this script so that it can call the same version of
662 A note about Exim versions 4.44 and 4.50
663 ----------------------------------------
665 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
666 changes of various kinds. As a consequence, a big documentation update was
667 needed. This delayed the release for rather longer than seemed good, especially
668 in the light of a couple of (minor) security issues. Therefore, the changes
669 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
670 release. So 4.44 and 4.50 are in effect two different branches that both start
673 I have left the 4.50 change log unchanged; it contains all the changes since
674 4.43. The change log for 4.44 is below; many of its items are identical to
675 those for 4.50. This seems to be the most sensible way to preserve the
676 historical information.
682 1. Minor wording change to the doc/README.SIEVE file.
684 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
685 computation of the current number of files was incorrect.
687 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
688 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
691 4. Give more explanation in the error message when the command for a transport
692 filter fails to execute.
694 5. There are several places where Exim runs a non-Exim command in a
695 subprocess. The SIGUSR1 signal should be disabled for these processes. This
696 was being done only for the command run by the queryprogram router. It is
697 now done for all such subprocesses. The other cases are: ${run, transport
698 filters, and the commands run by the lmtp and pipe transports.
700 6. Added CONFIGURE_GROUP build-time option.
702 7. Some older OS have a limit of 256 on the maximum number of file
703 descriptors. Exim was using setrlimit() to set 1000 as a large value
704 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
705 systems. I've change it so that if it can't get 1000, it tries for 256.
707 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
708 was an oversight, and furthermore, ever since the addition of extra
709 controls (e.g. 4.43/32), the checks on when to allow different forms of
710 "control" were broken. There should now be diagnostics for all cases when a
711 control that does not make sense is encountered.
713 9. Added the /retain_sender option to "control=submission".
715 10. $recipients is now available in the predata ACL (oversight).
717 11. Tidy the search cache before the fork to do a delivery from a message
718 received from the command line. Otherwise the child will trigger a lookup
719 failure and thereby defer the delivery if it tries to use (for example) a
720 cached ldap connection that the parent has called unbind on.
722 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
723 of $address_data from the recipient verification was clobbered by the
726 13. The value of address_data from a sender verification is now available in
727 $sender_address_data in subsequent conditions in the ACL statement.
729 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
731 15. Added a new option "connect=<time>" to callout options, to set a different
734 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
735 was its contents. (It was OK if the option was not defined at all.)
737 17. A "Completed" log line is now written for messages that are removed from
738 the spool by the -Mrm option.
740 18. New variables $sender_verify_failure and $recipient_verify_failure contain
741 information about exactly what failed.
743 19. Added -dd to debug only the daemon process.
745 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
746 handles timeouts, both on the server side and network timeouts. Renamed the
747 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
749 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
750 It is now set to "smtps".
752 22. $host_address is now set to the target address during the checking of
755 23. When checking ignore_target_hosts for an ipliteral router, no host name was
756 being passed; this would have caused $sender_host_name to have been used if
757 matching the list had actually called for a host name (not very likely,
758 since this list is usually IP addresses). A host name is now passed as
761 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
762 code that specifies a non-restarting handler (typically sigaction() in
763 modern systems) in an attempt to fix a rare and obscure crash bug.
765 25. Narrowed the window for a race in the daemon that could cause it to ignore
766 SIGCHLD signals. This is not a major problem, because they are used only to
767 wake it up if nothing else does.
769 26. A malformed maildirsize file could cause Exim to calculate negative values
770 for the mailbox size or file count. Odd effects could occur as a result.
771 The maildirsize information is now recalculated if the size or filecount
774 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
775 support for a long time. Removed HAVE_SYS_VFS_H.
777 28. Installed the latest version of exipick from John Jetmore.
779 29. In an address list, if the pattern was not a regular expression, an empty
780 subject address (from a bounce message) matched only if the pattern was an
781 empty string. Non-empty patterns were not even tested. This was the wrong
782 because it is perfectly reasonable to use an empty address as part of a
783 database query. An empty address is now tested by patterns that are
784 lookups. However, all the other forms of pattern expect the subject to
785 contain a local part and a domain, and therefore, for them, an empty
786 address still always fails if the pattern is not itself empty.
788 30. Exim went into a mad DNS loop when attempting to do a callout where the
789 host was specified on an smtp transport, and looking it up yielded more
792 31. Re-factored the code for checking spool and log partition space into a
793 function that finds that data and another that does the check. The former
794 is then used to implement four new variables: $spool_space, $log_space,
795 $spool_inodes, and $log_inodes.
797 32. The RFC2047 encoding function was originally intended for short strings
798 such as real names; it was not keeping to the 75-character limit for
799 encoded words that the RFC imposes. It now respects the limit, and
800 generates multiple encoded words if necessary. To be on the safe side, I
801 have increased the buffer size for the ${rfc2047: expansion operator from
804 33. It is now permitted to omit both strings after an "if" condition; if the
805 condition is true, the result is "true". As before, when the second string
806 is omitted, a false condition yields an empty string. This makes it less
807 cumbersome to write custom ACL and router conditions.
809 34. Failure to deliver a bounce message always caused it to be frozen, even if
810 there was an errors_to setting on the router. The errors_to setting is now
813 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
814 canonical form (fully expanded) before being placed in
815 $sender_host_address.
817 36. The table in the code that translates DNS record types into text (T_A to
818 "A" for instance) was missing entries for NS and CNAME. It is just possible
819 that this could have caused confusion if both these types were looked up
820 for the same domain, because the text type is used as part of Exim's
821 per-process caching. But the chance of anyone hitting this buglet seems
824 37. The dnsdb lookup has been extended in a number of ways.
826 (1) There is a new type, "zns", which walks up the domain tree until it
827 finds some nameserver records. It should be used with care.
829 (2) There is a new type, "mxh", which is like "mx" except that it returns
830 just the host names, not the priorities.
832 (3) It is now possible to give a list of domains (or IP addresses) to be
833 looked up. The behaviour when one of the lookups defers can be
834 controlled by a keyword.
836 (4) It is now possible to specify the separator character for use when
837 multiple records are returned.
839 38. The dnslists ACL condition has been extended: it is now possible to supply
840 a list of IP addresses and/or domains to be looked up in a particular DNS
843 39. Added log_selector=+queue_time_overall.
845 40. When running the queue in the test harness, wait just a tad after forking a
846 delivery process, to get repeatability of debugging output.
848 41. Include certificate and key file names in error message when GnuTLS fails
849 to set them up, because the GnuTLS error message doesn't include the name
850 of the failing file when there is a problem reading it.
852 42. Allow both -bf and -bF in the same test run.
854 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
856 44. The "Exiscan patch" is now merged into the mainline Exim source.
858 45. Sometimes the final signoff response after QUIT could fail to get
859 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
860 < 0 before doing a fflush(). This bug looks as though it goes back to the
861 introduction of TLS in release 3.20, but "sometimes" must have been rare
862 because the tests only now provoked it.
864 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
865 (this can affect the format of dates).
867 47. exim_tidydb, when checking for the continued existence of a message for
868 which it has found a message-specific retry record, was not finding
869 messages that were in split spool directories. Consequently, it was
870 deleting retry records that should have stayed in existence.
872 48. Steve fixed some bugs in eximstats.
874 49. The SPA authentication driver was not abandoning authentication and moving
875 on to the next authenticator when an expansion was forced to fail,
876 contradicting the general specification for all authenticators. Instead it
877 was generating a temporary error. It now behaves as specified.
879 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
880 (the order specifies the preference for clients). The order is now AES256,
881 AES128, 3DES, ARCFOUR128.
883 51. Small patch to Sieve code - explicitly set From: when generating an
886 52. Exim crashed if a remote delivery caused a very long error message to be
887 recorded - for instance if somebody sent an entire SpamAssassin report back
888 as a large number of 550 error lines. This bug was coincidentally fixed by
889 increasing the size of one of Exim's internal buffers (big_buffer) that
890 happened as part of the Exiscan merge. However, to be on the safe side, I
891 have made the code more robust (and fixed the comments that describe what
894 53. Now that there can be additional text after "Completed" in log lines (if
895 the queue_time_overall log selector is set), a one-byte patch to exigrep
896 was needed to allow it to recognize "Completed" as not the last thing in
899 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
900 patch that reportedly fixes this has been added. I am not expert enough to
901 create a test for it. This is what the patch creator wrote:
903 "I found a little strange behaviour of ldap code when working with
904 Windows 2003 AD Domain, where users was placed in more than one
905 Organization Units. When I tried to give exim partial DN, the exit code
906 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
907 But simultaneously result of request was absolutely normal ldap result,
908 so I produce this patch..."
910 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
911 REFERENCE, so I have modified the code to exclude the patch when that macro
914 55. Some experimental protocols are using DNS PTR records for new purposes. The
915 keys for these records are domain names, not reversed IP addresses. The
916 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
917 leaves it alone. Component reversal etc. now happens only for IP addresses.
920 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
922 57. Double the size of the debug message buffer (to 2048) so that more of very
923 long debug lines gets shown.
925 58. The exicyclog utility now does better if the number of log files to keep
926 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
928 59. Two changes related to the smtp_active_hostname option:
930 (1) $smtp_active_hostname is now available as a variable.
931 (2) The default for smtp_banner uses $smtp_active_hostname instead
932 of $primary_hostname.
934 60. The host_aton() function is supposed to be passed a string that is known
935 to be a valid IP address. However, in the case of IPv6 addresses, it was
936 not checking this. This is a hostage to fortune. Exim now panics and dies
937 if the condition is not met. A case was found where this could be provoked
938 from a dnsdb PTR lookup with an IPv6 address that had more than 8
939 components; fortuitously, this particular loophole had already been fixed
940 by change 4.50/55 above.
942 If there are any other similar loopholes, the new check in host_aton()
943 itself should stop them being exploited. The report I received stated that
944 data on the command line could provoke the exploit when Exim was running as
945 exim, but did not say which command line option was involved. All I could
946 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
950 61. There was a buffer overflow vulnerability in the SPA authentication code
951 (which came originally from the Samba project). I have added a test to the
952 spa_base64_to_bits() function which I hope fixes it.
955 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
956 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
958 63. The daemon start-up calls getloadavg() while still root for those OS that
959 need the first call to be done as root, but it missed one case: when
960 deliver_queue_load_max is set with deliver_drop_privilege. This is
961 necessary for the benefit of the queue runner, because there is no re-exec
962 when deliver_drop_privilege is set.
964 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
967 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
968 was in use, was not putting the data itself into the right store pool;
969 consequently, it could be overwritten for a subsequent message in the same
970 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
973 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
975 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
976 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
977 the function were treating the return as a boolean value, which happened to
978 work because 0=false and not-0=true, but is not correct code.
980 68. The host_aton() function was not handling scoped IPv6 addresses (those
981 with, for example, "%eth0" on the end) correctly.
983 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
984 negated items (that is, ~something) in unsigned ints. Some compilers
985 apparently mutter when there is no cast.
987 70. If an address verification called from an ACL failed, and did not produce a
988 user-specific message (i.e. there was only a "system" message), nothing was
989 put in $acl_verify_message. In this situation, it now puts the system
992 71. Change 4.23/11 added synchronization checking at the start of an SMTP
993 session; change 4.31/43 added the unwanted input to the log line - except
994 that it did not do this in the start of session case. It now does.
996 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
997 This is wrong and can cause the other end to generate a synchronization
998 error if it is another Exim or anything else that does the synchronization
999 check. A QUIT command is no longer sent after a timeout.
1001 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1002 during host lookups.
1004 74. The defer_ok option of callout verification was not working if it was used
1005 when verifying addresses in header lines, that is, for this case:
1007 verify = header_sender/callout=defer_ok
1009 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1010 those file descriptors could be used for SMTP connections. If anything
1011 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1012 could be sent to the SMTP client, causing chaos. The daemon now opens
1013 stdin, stdout, and stderr to /dev/null when it puts itself into the
1016 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1017 log by default. The user can override this with suitable Perl magic.
1019 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1020 the log message when discard triggers, was not working for the DATA ACL or
1021 for the non-SMTP ACL.
1023 78. Error message wording change in sieve.c.
1025 79. If smtp_accept_max_per_host was set, the number of connections could be
1026 restricted to fewer than expected, because the daemon was trying to set up
1027 a new connection before checking whether the processes handling previous
1028 connections had finished. The check for completed processes is now done
1029 earlier. On busy systems, this bug wouldn't be noticed because something
1030 else would have woken the daemon, and it would have reaped the completed
1033 80. If a message was submitted locally by a user whose login name contained one
1034 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1035 It caused a spool format error. I have fixed the spool reading code. A
1036 related problem was that the "from" clause in the Received: line became
1037 illegal because of the space(s). It is now covered by ${quote_local_part.
1039 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1042 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1044 83. Patch from Sieve maintainer to fix unterminated string problem in
1045 "vacation" handling.
1047 84. Some minor changes to the Linux configuration files to help with other
1048 OS variants using glibc.
1050 85. One more patch for Sieve to update vacation handling to latest spec.
1053 ----------------------------------------------------
1054 See the note above about the 4.44 and 4.50 releases.
1055 ----------------------------------------------------
1061 1. Change 4.43/35 introduced a bug that caused file counts to be
1062 incorrectly computed when quota_filecount was set in an appendfile
1065 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1066 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1069 3. Give more explanation in the error message when the command for a transport
1070 filter fails to execute.
1072 4. There are several places where Exim runs a non-Exim command in a
1073 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1074 was being done only for the command run by the queryprogram router. It is
1075 now done for all such subprocesses. The other cases are: ${run, transport
1076 filters, and the commands run by the lmtp and pipe transports.
1078 5. Some older OS have a limit of 256 on the maximum number of file
1079 descriptors. Exim was using setrlimit() to set 1000 as a large value
1080 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1081 systems. I've change it so that if it can't get 1000, it tries for 256.
1083 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1084 was an oversight, and furthermore, ever since the addition of extra
1085 controls (e.g. 4.43/32), the checks on when to allow different forms of
1086 "control" were broken. There should now be diagnostics for all cases when a
1087 control that does not make sense is encountered.
1089 7. $recipients is now available in the predata ACL (oversight).
1091 8. Tidy the search cache before the fork to do a delivery from a message
1092 received from the command line. Otherwise the child will trigger a lookup
1093 failure and thereby defer the delivery if it tries to use (for example) a
1094 cached ldap connection that the parent has called unbind on.
1096 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1097 of $address_data from the recipient verification was clobbered by the
1098 sender verification.
1100 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1101 was its contents. (It was OK if the option was not defined at all.)
1103 11. A "Completed" log line is now written for messages that are removed from
1104 the spool by the -Mrm option.
1106 12. $host_address is now set to the target address during the checking of
1107 ignore_target_hosts.
1109 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1110 being passed; this would have caused $sender_host_name to have been used if
1111 matching the list had actually called for a host name (not very likely,
1112 since this list is usually IP addresses). A host name is now passed as
1115 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1116 code that specifies a non-restarting handler (typically sigaction() in
1117 modern systems) in an attempt to fix a rare and obscure crash bug.
1119 15. Narrowed the window for a race in the daemon that could cause it to ignore
1120 SIGCHLD signals. This is not a major problem, because they are used only to
1121 wake it up if nothing else does.
1123 16. A malformed maildirsize file could cause Exim to calculate negative values
1124 for the mailbox size or file count. Odd effects could occur as a result.
1125 The maildirsize information is now recalculated if the size or filecount
1128 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1129 support for a long time. Removed HAVE_SYS_VFS_H.
1131 18. Updated exipick to current release from John Jetmore.
1133 19. Allow an empty sender to be matched against a lookup in an address list.
1134 Previously the only cases considered were a regular expression, or an
1137 20. Exim went into a mad DNS lookup loop when doing a callout where the
1138 host was specified on the transport, if the DNS lookup yielded more than
1141 21. The RFC2047 encoding function was originally intended for short strings
1142 such as real names; it was not keeping to the 75-character limit for
1143 encoded words that the RFC imposes. It now respects the limit, and
1144 generates multiple encoded words if necessary. To be on the safe side, I
1145 have increased the buffer size for the ${rfc2047: expansion operator from
1148 22. Failure to deliver a bounce message always caused it to be frozen, even if
1149 there was an errors_to setting on the router. The errors_to setting is now
1152 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1153 canonical form (fully expanded) before being placed in
1154 $sender_host_address.
1156 24. Updated eximstats to version 1.33
1158 25. Include certificate and key file names in error message when GnuTLS fails
1159 to set them up, because the GnuTLS error message doesn't include the name
1160 of the failing file when there is a problem reading it.
1162 26. Expand error message when OpenSSL has problems setting up cert/key files.
1165 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1166 (this can affect the format of dates).
1168 28. exim_tidydb, when checking for the continued existence of a message for
1169 which it has found a message-specific retry record, was not finding
1170 messages that were in split spool directories. Consequently, it was
1171 deleting retry records that should have stayed in existence.
1173 29. eximstats updated to version 1.35
1174 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1175 1.35 - bugfix such that pie charts by volume are generated correctly
1177 30. The SPA authentication driver was not abandoning authentication and moving
1178 on to the next authenticator when an expansion was forced to fail,
1179 contradicting the general specification for all authenticators. Instead it
1180 was generating a temporary error. It now behaves as specified.
1182 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1183 (the order specifies the preference for clients). The order is now AES256,
1184 AES128, 3DES, ARCFOUR128.
1186 31. Small patch to Sieve code - explicitly set From: when generating an
1189 32. Exim crashed if a remote delivery caused a very long error message to be
1190 recorded - for instance if somebody sent an entire SpamAssassin report back
1191 as a large number of 550 error lines. This bug was coincidentally fixed by
1192 increasing the size of one of Exim's internal buffers (big_buffer) that
1193 happened as part of the Exiscan merge. However, to be on the safe side, I
1194 have made the code more robust (and fixed the comments that describe what
1197 33. Some experimental protocols are using DNS PTR records for new purposes. The
1198 keys for these records are domain names, not reversed IP addresses. The
1199 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1200 leaves it alone. Component reversal etc. now happens only for IP addresses.
1203 34. The host_aton() function is supposed to be passed a string that is known
1204 to be a valid IP address. However, in the case of IPv6 addresses, it was
1205 not checking this. This is a hostage to fortune. Exim now panics and dies
1206 if the condition is not met. A case was found where this could be provoked
1207 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1208 components; fortuitously, this particular loophole had already been fixed
1209 by change 4.50/55 or 4.44/33 above.
1211 If there are any other similar loopholes, the new check in host_aton()
1212 itself should stop them being exploited. The report I received stated that
1213 data on the command line could provoke the exploit when Exim was running as
1214 exim, but did not say which command line option was involved. All I could
1215 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1216 running as the user.
1219 35. There was a buffer overflow vulnerability in the SPA authentication code
1220 (which came originally from the Samba project). I have added a test to the
1221 spa_base64_to_bits() function which I hope fixes it.
1224 36. The daemon start-up calls getloadavg() while still root for those OS that
1225 need the first call to be done as root, but it missed one case: when
1226 deliver_queue_load_max is set with deliver_drop_privilege. This is
1227 necessary for the benefit of the queue runner, because there is no re-exec
1228 when deliver_drop_privilege is set.
1230 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1231 was in use, was not putting the data itself into the right store pool;
1232 consequently, it could be overwritten for a subsequent message in the same
1233 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1236 38. Sometimes the final signoff response after QUIT could fail to get
1237 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1238 < 0 before doing a fflush(). This bug looks as though it goes back to the
1239 introduction of TLS in release 3.20, but "sometimes" must have been rare
1240 because the tests only now provoked it.
1246 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1247 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1248 Now it returns an integer. A number of calls were still expecting a T/F
1249 return. Fortuitously, in all cases, the tests worked in OK situations,
1250 which is the norm. However, things would have gone wrong on any write
1251 failures on the smtp file descriptor. This function is used when sending
1252 messages over SMTP and also when doing verify callouts.
1254 2. When Exim is called to do synchronous delivery of a locally submitted
1255 message (the -odf or -odi options), it no longer closes stderr before doing
1258 3. Implemented the mua_wrapper option.
1260 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1262 5. Implemented the functions header_remove(), header_testname(),
1263 header_add_at_position(), and receive_remove_recipient(), and exported them
1266 6. If an ACL "warn" statement specified the addition of headers, Exim already
1267 inserted X-ACL-Warn: at the start if there was no header name. However, it
1268 was not making this test for the second and subsequent header lines if
1269 there were newlines in the string. This meant that an invalid header could
1270 be inserted if Exim was badly configured.
1272 7. Allow an ACL "warn" statement to add header lines at the start or after all
1273 the Received: headers, as well as at the end.
1275 8. Added the rcpt_4xx retry error code.
1277 9. Added postmaster_mailfrom=xxx to callout verification option.
1279 10. Added mailfrom=xxxx to the callout verification option, for verify=
1282 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1283 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1285 12. Inserted some casts to stop certain compilers warning when using pointer
1286 differences as field lengths or precisions in printf-type calls (mostly
1287 affecting debugging statements).
1289 13. Added optional readline() support for -be (dynamically loaded).
1291 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1292 same clock tick as a message's arrival, so that its received time was the
1293 same as the "first fail" time on the retry record, and that message
1294 remained on the queue past the ultimate address timeout, every queue runner
1295 would try a delivery (because it was past the ultimate address timeout) but
1296 after another failure, the ultimate address timeout, which should have then
1297 bounced the address, did not kick in. This was a "< instead of <=" error;
1298 in most cases the first failure would have been in the next clock tick
1299 after the received time, and all would be well.
1301 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1302 being recognized when the domain list was tested by the match_domain
1303 condition in an expansion string.
1305 16. Added the ${str2b64: operator.
1307 17. Exim was always calling setrlimit() to set a large limit for the number of
1308 processes, without checking whether the existing limit was already
1309 adequate. (It did check for the limit on file descriptors.) Furthermore,
1310 errors from getrlimit() and setrlimit() were being ignored. Now they are
1311 logged to the main and panic logs, but Exim does carry on, to try to do its
1312 job under whatever limits there are.
1314 18. Imported PCRE 5.0.
1316 19. Trivial typo in log message " temporarily refused connection" (the leading
1319 20. If the log selector return_path_on_delivery was set and an address was
1320 redirected to /dev/null, the delivery process crashed because it assumed
1321 that a return path would always be set for a "successful" delivery. In this
1322 case, the whole delivery is bypassed as an optimization, and therefore no
1325 21. Internal re-arrangement: the function for sending a challenge and reading
1326 a response while authentication was assuming a zero-terminated challenge
1327 string. It's now changed to take a pointer and a length, to allow for
1328 binary data in such strings.
1330 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1332 23. Exim was not respecting finduser_retries when seeking the login of the
1333 uid under which it was called; it was always trying 10 times. (The default
1334 setting of finduser_retries is zero.) Also, it was sleeping after the final
1335 failure, which is pointless.
1337 24. Implemented tls_on_connect_ports.
1339 25. Implemented acl_smtp_predata.
1341 26. If the domain in control=submission is set empty, Exim assumes that the
1342 authenticated id is a complete email address when it generates From: or
1343 Sender: header lines.
1345 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1346 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1347 chown and chgrp in /bin and hostname in /usr/bin.
1349 28. Exim was keeping the "process log" file open after each use, just as it
1350 does for the main log. This opens the possibility of it remaining open for
1351 long periods when the USR1 signal hits a daemon. Occasional processlog
1352 errors were reported, that could have been caused by this. Anyway, it seems
1353 much more sensible not to leave this file open at all, so that is what now
1356 29. The long-running daemon process does not normally write to the log once it
1357 has entered its main loop, and it closes the log before doing so. This is
1358 so that log files can straightforwardly be renamed and moved. However,
1359 there are a couple of unusual error situations where the daemon does write
1360 log entries, and I had neglected to close the log afterwards.
1362 30. The text of an SMTP error response that was received during a remote
1363 delivery was being truncated at 512 bytes. This is too short for some of
1364 the long messages that one sometimes sees. I've increased the limit to
1367 31. It is now possible to make retry rules that apply only when a message has a
1368 specific sender, in particular, an empty sender.
1370 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1371 it possible to be selective about when SMTP synchronization is enforced.
1373 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1375 32. Implemented hosts_connection_nolog.
1377 33. Added an ACL for QUIT.
1379 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1382 35. Added mailbox_size and mailbox_filecount to appendfile.
1384 36. Added control = no_multiline_responses to ACLs.
1386 37. There was a bug in the logic of the code that waits for the clock to tick
1387 in the case where the clock went backwards by a substantial amount such
1388 that the microsecond fraction of "now" was more than the microsecond
1389 fraction of "then" (but the whole seconds number was less).
1391 38. Added support for the libradius Radius client library this is found on
1392 FreeBSD (previously only the radiusclient library was supported).
1398 1. When certain lookups returned multiple values in the form name=value, the
1399 quoting of the values was not always being done properly. Specifically:
1400 (a) If the value started with a double quote, but contained no whitespace,
1402 (b) If the value contained whitespace other than a space character (i.e.
1403 tabs or newlines or carriage returns) it was not quoted.
1404 This fix has been applied to the mysql and pgsql lookups by writing a
1405 separate quoting function and calling it from the lookup code. The fix
1406 should probably also be applied to nisplus, ibase and oracle lookups, but
1407 since I cannot test any of those, I have not disturbed their existing code.
1409 2. A hit in the callout cache for a specific address caused a log line with no
1410 reason for rejecting RCPT. Now it says "Previous (cached) callout
1411 verification failure".
1413 3. There was an off-by-one bug in the queryprogram router. An over-long
1414 return line was truncated at 256 instead of 255 characters, thereby
1415 overflowing its buffer with the terminating zero. As well as fixing this, I
1416 have increased the buffer size to 1024 (and made a note to document this).
1418 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1419 when Exim is waiting for an SMTP response from a remote server, Exim
1420 restarts its select() call on the socket, thereby resetting its timeout.
1421 This is not a problem when such interrupts are rare. Somebody set up a cron
1422 job to run exiwhat every 2 minutes, which is less than the normal select()
1423 timeout (5 or 10 minutes). This meant that the select() timeout never
1424 kicked in because it was always reset. I have fixed this by comparing the
1425 time when an interrupt arrives with the time at the start of the first call
1426 to select(). If more time than the timeout has elapsed, the interrupt is
1427 treated as a timeout.
1429 5. Some internal re-factoring in preparation for the addition of Sieve
1430 extensions (by MH). In particular, the "personal" test is moved to a
1431 separate function, and given an option for scanning Cc: and Bcc: (which is
1432 not set for Exim filters).
1434 6. When Exim created an email address using the login of the caller as the
1435 local part (e.g. when creating a From: or Sender: header line), it was not
1436 quoting the local part when it contained special characters such as @.
1438 7. Installed new OpenBSD configuration files.
1440 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1441 try to make them clearer.
1443 9. Callout options, other than the timeout value, were being ignored when
1444 verifying sender addresses in header lines. For example, when using
1446 verify = header_sender/callout=no_cache
1448 the cache was (incorrectly) being used.
1450 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1451 only the Cygwin environment.
1453 11. When return_path_on_delivery was set as a log selector, if different remote
1454 addresses in the same message used different return paths and parallel
1455 remote delivery occurred, the wrong values would sometimes be logged.
1456 (Whenever a remote delivery process finished, the return path value from
1457 the most recently started remote delivery process was logged.)
1459 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1460 lines when AUTH and/or TLS are in use. This is the "received protocol"
1461 field. Exim used to use "asmtp" for authenticated SMTP, without any
1462 indication (in the protocol name) for TLS use. Now it follows the RFC and
1463 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1464 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1465 names appear in log lines as well as in Received: header lines.
1467 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1468 extensions, and comparison tests, and to fix some bugs.
1470 14. Changes to the "personal" filter test:
1472 (1) The test was buggy in that it was just doing the equivalent of
1473 "contains" tests on header lines. For example, if a user's address was
1474 anne@some.where, the "personal" test would incorrectly be true for
1476 To: susanne@some.where
1478 This test is now done by extracting each address from the header in turn,
1479 and checking the entire address. Other tests that are part of "personal"
1480 are now done using regular expressions (for example, to check local parts
1481 of addresses in From: header lines).
1483 (2) The list of non-personal local parts in From: addresses has been
1484 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1485 taken from the Sieve specification recommendations.
1487 (3) If the message contains any header line starting with "List-" it is
1488 treated as non-personal.
1490 (4) The test for "circular" in the Subject: header line has been removed
1491 because it now seems ill-conceived.
1493 15. Minor typos in src/EDITME comments corrected.
1495 16. Installed latest exipick from John Jetmore.
1497 17. If headers_add on a router specified a text string that was too long for
1498 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1499 of string_sprintf() is now avoided.
1501 18. $message_body_size was not set (it was always zero) when running the DATA
1502 ACL and the local_scan() function.
1504 19. For the "mail" command in an Exim filter, no default was being set for
1505 the once_repeat time, causing a random time value to be used if "once" was
1506 specified. (If the value happened to be <= 0, no repeat happened.) The
1507 default is now 0s, meaning "never repeat". The "vacation" command was OK
1508 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1509 (I found it when inspecting the code).
1511 20. There is now an overall timeout for performing a callout verification. It
1512 defaults to 4 times the callout timeout, which applies to individual SMTP
1513 commands during the callout. The overall timeout applies when there is more
1514 than one host that can be tried. The timeout is checked before trying the
1515 next host. This prevents very long delays if there are a large number of
1516 hosts and all are timing out (e.g. when the network connections are timing
1517 out). The value of the overall timeout can be changed by specifying an
1518 additional sub-option for "callout", called "maxwait". For example:
1520 verify = sender/callout=5s,maxwait=20s
1522 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1523 to the end before writing, but this should make it even safer).
1525 22. Exim was forgetting that it had advertised PIPELINING for the second and
1526 subsequent messages on an SMTP connection. It was also not resetting its
1527 memory on STARTTLS and an internal HELO.
1529 23. When Exim logs an SMTP synchronization error within a session, it now
1530 records whether PIPELINING has been advertised or not.
1532 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1533 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1534 rather than long int.
1536 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1538 26. Added the never_mail option to autoreply.
1544 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1545 crash if the getsockname() call failed; this can happen if a connection is
1546 closed very soon after it is established. The problem was simply in the
1547 order in which certain operations were done, causing Exim to try to write
1548 to the SMTP stream before it had set up the file descriptor. The bug has
1549 been fixed by making things happen in the correct order.
1555 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1556 before the connection was closed, thus losing the rejection response.
1558 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1559 some early Solaris releases, but causes trouble in current releases where
1560 socklen_t is defined.
1562 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1565 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1568 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1569 that releasing the top of it at the end releases what was used for sub-
1570 expansions (unless the block got too big). However, discard this block if
1571 the first thing is a variable or header, so that we can use its block when
1572 it is dynamic (useful for very large $message_headers, for example).
1574 6. Lookups now cache *every* query, not just the most recent. A new, separate
1575 store pool is used for this. It can be recovered when all lookup caches are
1576 flushed. Lookups now release memory at the end of their result strings.
1577 This has involved some general refactoring of the lookup sources.
1579 7. Some code has been added to the store_xxx() functions to reduce the amount
1580 of flapping under certain conditions.
1582 8. log_incoming_interface used to affect only the <= reception log lines. Now
1583 it causes the local interface and port to be added to several more SMTP log
1584 lines, for example "SMTP connection from", and rejection lines.
1586 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1588 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1590 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1591 could be overwritten at the end of the current message (or the start of a
1592 new message if it was set in a HELO ACL). The value is now preserved for
1593 the duration of the SMTP connection.
1595 12. If a transport had a headers_rewrite setting, and a matching header line
1596 contained an unqualified address, that address was qualified, even if it
1597 did not match any rewriting rules. The underlying bug was that the values
1598 of the flags that permit the existence of unqualified sender and recipient
1599 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1600 non-local messages, and by -bnq for local messages) were not being
1601 preserved with the message after it was received.
1603 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1604 "next input=" as part of the text comprising the host identity instead of
1605 the correct text. The code was using the same buffer for two different
1606 strings. However, depending on which order the printing function evaluated
1607 its arguments, the bug did not always show up. Under Linux, for example, my
1608 test suite worked just fine.
1610 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1611 doesn't work with some older versions of Perl. It has been changed to "my",
1612 which in any case is probably the better facility to use.
1614 15. A really picky compiler found some instances of statements for creating
1615 error messages that either had too many or two few arguments for the format
1618 16. The size of the buffer for calls to the DNS resolver has been increased
1619 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1620 for addresses that have a lot of PTR records. This alleviates a problem; it
1621 does not fully solve it.
1623 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1624 buffer now truncates the list and logs the incident, which is the same
1625 action as happens when Exim is looking up a host name and its aliases.
1626 Previously in this situation something unpredictable would happen;
1627 sometimes it was "internal error: store_reset failed".
1629 18. If a server dropped the connection unexpectedly when an Exim client was
1630 using GnuTLS and trying to read a response, the client delivery process
1631 crashed while trying to generate an error log message.
1633 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1634 string, for example:
1636 warn message = H1: something\nH2: something
1638 the text was added as a single header line from Exim's point of view
1639 though it ended up OK in the delivered message. However, searching for the
1640 second and subsequent header lines using $h_h2: did not work. This has been
1641 fixed. Similarly, if a system filter added multiple headers in this way,
1642 the routers could not see them.
1644 20. Expanded the error message when iplsearch is called with an invalid key to
1645 suggest using net-iplsearch in a host list.
1647 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1648 ACLs are no longer actually imposed (and a message to that effect is
1651 22. If a "gecos" field in a passwd entry contained escaped characters, in
1652 particular, if it contained a \" sequence, Exim got it wrong when building
1653 a From: or a Sender: header from that name. A second bug also caused
1654 incorrect handling when an unquoted " was present following a character
1655 that needed quoting.
1657 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1658 was not being matched caselessly.
1660 24. Arranged for all hyphens in the exim.8 source to be escaped with
1663 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1664 properly. Recipient callout cache records were still being keyed to include
1665 the sender, even when use_sender was set false. This led to far more
1666 callouts that were necessary. The sender is no longer included in the key
1667 when use_sender is false.
1669 26. Added "control = submission" modifier to ACLs.
1671 27. Added the ${base62d: operator to decode base 62 numbers.
1673 28. dnsdb lookups can now access SRV records.
1675 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1676 the configuration file.
1678 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1679 (-v) mode. This makes the output for a verbose queue run more intelligible.
1681 31. Added a use_postmaster feature to recipient callouts.
1683 32. Added the $body_zerocount variable, containing the number of binary zero
1684 bytes in the message body.
1686 33. The time of last modification of the "new" subdirectory is now used as the
1687 "mailbox time last read" when there is a quota error for a maildir
1690 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1692 35. Added +ignore_unknown as a special item in host lists.
1694 36. Code for decoding IPv6 addresses in host lists is now included, even if
1695 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1696 address was recognized as an IP address, but was then not correctly decoded
1697 into binary, causing unexpected and incorrect effects when compared with
1704 1. Very minor rewording of debugging text in manualroute to say "list of
1705 hosts" instead of "hostlist".
1707 2. If verify=header_syntax was set, and a header line with an unqualified
1708 address (no domain) and a large number of spaces between the end of the
1709 name and the colon was received, the reception process suffered a buffer
1710 overflow, and (when I tested it) crashed. This was caused by some obsolete
1711 code that should have been removed. The fix is to remove it!
1713 3. When running in the test harness, delay a bit after writing a bounce
1714 message to get a bit more predictability in the log output.
1716 4. Added a call to search_tidyup() just before forking a reception process. In
1717 theory, someone could use a lookup in the expansion of smtp_accept_max_
1718 per_host which, without the tidyup, could leave open a database connection.
1720 5. Added the variables $recipient_data and $sender_data which get set from a
1721 lookup success in an ACL "recipients" or "senders" condition, or a router
1722 "senders" option, similar to $domain_data and $local_part_data.
1724 6. Moved the writing of debug_print from before to after the "senders" test
1727 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1728 problems for message scanning, either using a data ACL, or using
1729 local_scan() because the Received: header was not generated till after they
1730 were called (in order to set the time as the time of reception completion).
1731 I have revised the way this works. The header is now generated after the
1732 body is received, but before the ACL or local_scan() are called. After they
1733 are run, the timestamp in the header is updated.
1739 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1740 before starting a queue runner without re-exec. This happened only when
1741 deliver_drop_privilege was set or when the Exim user was set to root. The
1742 effect of the bug was that timeouts during subsequent deliveries caused
1743 crashes instead of being properly handled. The handler is now left at its
1744 default (and expected) setting.
1746 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1747 message, again when deliver_drop_privilege is set or Exim is run as root.
1748 The bug described in (1) was not present in this case, but the tidying up
1749 of the other signals was missing. I have made the two cases consistent.
1751 3. The ignore_target_hosts setting on a manualroute router was being ignored
1752 for hosts that were looked up using the /MX notation.
1754 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1757 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1758 operated on the sender address. After changing the $sender_address to <>
1759 for the sender address verify, Exim was re-instated it as the original
1760 (before rewriting) address, but remembering that it had rewritten it, so it
1761 wasn't rewriting it again. This bug also had the effect of breaking the
1762 sender address verification caching when the sender address was rewritten.
1764 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1765 This has been changed so that if the ip literal address matches
1766 ignore_target_hosts, the router declines.
1768 7. Added expansion conditions match_domain, match_address, and match_local_
1769 part (NOT match_host).
1771 8. The placeholder for the Received: header didn't have a length field set.
1773 9. Added code to Exim itself and to exim_lock to test for a specific race
1774 condition that could lead to file corruption when using MBX delivery. The
1775 issue is with the lockfile that is created in /tmp. If this file is removed
1776 after a process has opened it but before that process has acquired a lock,
1777 there is the potential for a second process to recreate the file and also
1778 acquire a lock. This could lead to two Exim processes writing to the file
1779 at the same time. The added code performs the same test as UW imapd; it
1780 checks after acquiring the lock that its file descriptor still refers to
1781 the same named file.
1783 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1784 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1787 11. Added the smtp_active_hostname option. If used, this will typically be made
1788 to depend on the incoming interface address. Because $interface_address is
1789 not set up until the daemon has forked a reception process, error responses
1790 that can happen earlier (such as "too many connections") no longer contain
1793 12. If an expansion in a condition on a "warn" statement fails because a lookup
1794 defers, the "warn" statement is abandoned, and the next ACL statement is
1795 processed. Previously this caused the whole ACL to be aborted.
1797 13. Added the iplsearch lookup type.
1799 14. Added ident_timeout as a log selector.
1801 15. Added tls_certificate_verified as a log selector.
1803 16. Added a global option tls_require_ciphers (compare the smtp transport
1804 option of the same name). This controls incoming TLS connections.
1806 17. I finally figured out how to make tls_require_ciphers do a similar thing
1807 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1808 before starting the TLS session.
1810 18. Tabs are now shown as \t in -bP output.
1812 19. If the log selector return_path_on_delivery was set, Exim crashed when
1813 bouncing a message because it had too many Received: header lines.
1815 20. If two routers both had headers_remove settings, and the first one included
1816 a superfluous trailing colon, the final name in the first list and the
1817 first name in the second list were incorrectly joined into one item (with a
1818 colon in the middle).
1824 1. Added -C and -D options to the exinext utility, mainly to make it easier
1825 to include in the automated testing, but these could be helpful when
1826 multiple configurations are in use.
1828 2. The exinext utility was not formatting the output nicely when there was
1829 an alternate port involved in the retry record key, nor when there was a
1830 message id as well (for retries that were specific to a specific message
1831 and a specific host). It was also confused by IPv6 addresses, because of
1832 the additional colons they contain. I have fixed the IPv4 problem, and
1833 patched it up to do a reasonable job for IPv6.
1835 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1836 delivery, the log line now contains "pipelined" if PIPELINING was used.
1838 4. An SMTP transport process used to panic and die if the bind() call to set
1839 an explicit outgoing interface failed. This has been changed; it is now
1840 treated in the same way as a connect() failure.
1842 5. A reference to $sender_host_name in the part of a conditional expansion
1843 that was being skipped was still causing a DNS lookup. This no longer
1846 6. The def: expansion condition was not recognizing references to header lines
1847 that used bh_ and bheader_.
1849 7. Added the _cache feature to named lists.
1851 8. The code for checking quota_filecount in the appendfile transport was
1852 allowing one more file than it should have been.
1854 9. For compatibility with Sendmail, the command line option
1862 and sets the incoming protocol and host name (for trusted callers). The
1863 host name and its colon can be omitted when only the protocol is to be set.
1864 Note the Exim already has two private options, -pd and -ps, that refer to
1865 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1866 "s", but I don't think that's a major issue.
1868 10. A number of refactoring changes to the code, none of which should affect
1871 (a) The number of logging options was getting close to filling up the
1872 32-bit word that was used as a bit map. I have split them into two classes:
1873 those that are passed in the argument to log_write(), and those that are
1874 only ever tested independently outside of that function. These are now in
1875 separate 32-bit words, so there is plenty of room for expansion again.
1876 There is no change in the user interface or the logging behaviour.
1878 (b) When building, for example, log lines, the code previously used a
1879 macro that called string_cat() twice, in order to add two strings. This is
1880 not really sufficiently general. Furthermore, there was one instance where
1881 it was actually wrong because one of the argument was used twice, and in
1882 one call a function was used. (As it happened, calling the function twice
1883 did not affect the overall behaviour.) The macro has been replaced by a
1884 function that can join an arbitrary number of extra strings onto a growing
1887 (c) The code for expansion conditions now uses a table and a binary chop
1888 instead of a serial search (which was left over from when there were very
1889 few conditions). Also, it now recognizes conditions like "pam" even when
1890 the relevant support is not compiled in: a suitably worded error message is
1891 given if an attempt is made to use such a condition.
1893 11. Added ${time_interval:xxxxx}.
1895 12. A bug was causing one of the ddress fields not to be passed back correctly
1896 from remote delivery subprocesses. The field in question was not being
1897 subsequently used, so this caused to problems in practice.
1899 13. Added new log selectors queue_time and deliver_time.
1901 14. Might have fixed a bug in maildirsizefile handling that threw up
1902 "unexpected character" debug warnings, and recalculated the data
1903 unnecessarily. In any case, I expanded the warning message to give more
1906 15. Added the message "Restricted characters in address" to the statements in
1907 the default ACL that block characters like @ and % in local parts.
1909 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1910 Three changes have been made:
1912 (a) There was a serious bug; a negative response to MAIL caused the whole
1913 recipient domain to be cached as invalid, thereby blocking all messages
1914 to all local parts at the same domain, from all senders. This bug has
1915 been fixed. The domain is no longer cached after a negative response to
1916 MAIL if the sender used is not empty.
1918 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1921 (c) A new callout option, "use_sender" has been added for people who want
1922 the modified behaviour.
1928 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1931 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1932 indeed breaks things for older releases.
1934 3. Added additional logging to the case where there is a problem reading data
1935 from a filter that is running in a subprocess using a pipe, in order to
1936 try to track down a specific problem.
1938 4. Testing facility fudge: when running in the test harness and attempting
1939 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1940 getting "No route to host". Convert this to a timeout.
1942 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1945 6. Some OS don't have socklen_t but use size_t instead. This affects the
1946 fifth argument of getsockopt() amongst other things. This is now
1947 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1948 can be set for individual OS. I have set it for SunOS5, OSF1, and
1949 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1950 some earlier ones do not.
1952 7. Change 4.30/15 was not doing the test caselessly.
1954 8. The standard form for an IPv6 address literal was being rejected by address
1955 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1956 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1957 this, as well as the form without the "IPv6" on the front (but only when
1958 address literals are enabled, of course).
1960 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1962 10. Exim crashed if a message with an empty sender address specified by -f
1963 encountered a router with an errors_to setting. This could be provoked only
1964 by a command such as
1968 where an empty string was supplied; "<>" did not hit this bug.
1970 11. Installed PCRE release 4.5.
1972 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1973 remained set. It is now erased.
1975 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1976 times from message ids (which are base 36 rather than the normal 62).
1978 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1979 were being counted as actual protocol errors, and logged if the log
1980 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1981 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1982 and DATA following a set of rejected RCPTs do not count as protocol errors.
1983 In other words, Exim assumes they were pipelined, though this may not
1984 actually be the case. Of course, in all cases the client gets an
1985 appropriate error code.
1987 15. If a lookup fails in an ACL condition, a message about the failure may
1988 be available; it is used if testing the ACL cannot continue, because most
1989 such messages specify what the cause of the deferral is. However, some
1990 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1991 that caused an old message to be retained and used if a later statement
1992 caused a defer, replacing the real cause of the deferral.
1994 16. If an IP address had so many PTR records that the DNS lookup buffer
1995 was not large enough to hold them, Exim could crash while trying to process
1996 the truncated data. It now detects and logs this case.
1998 17. Further to 4.21/58, another change has been made: if (and only if) the
1999 first line of a message (the first header line) ends with CRLF, a bare LF
2000 in a subsequent header line has a space inserted after it, so as not to
2001 terminate the header.
2003 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2004 unnecessarily, used atoi() instead of strtol(), and didn't check the
2005 termination when getting file sizes from file names by regex.
2007 19. Completely re-implemented the support for maildirsize files, in the light
2008 of a number of problems with the previous contributed implementation
2009 (4.30/29). In particular:
2011 . If the quota is zero, the maildirsize file is maintained, but no quota is
2014 . If the maildir directory does not exist, it is created before any attempt
2015 to write a maildirsize file.
2017 . The quota value in the file is just a cache; if the quota is changed in
2018 the transport, the new value overrides.
2020 . A regular expression is available for excluding directories from the
2023 20. The autoreply transport checks the characters in options that define the
2024 message's headers; it allows continued headers, but it was checking with
2025 isspace() after an embedded newline instead of explicitly looking for a
2028 21. If all the "regular" hosts to which an address was routed had passed their
2029 expiry times, and had not reached their retry times, the address was
2030 bounced, even if fallback hosts were defined. Now Exim should go on to try
2033 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2034 equivalent code in the SMTP transport. Some hosts send humungous responses
2035 to HELO/EHLO, more than 1024 it seems.
2037 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2038 gives compiler warnings in some environments. I've now done it "properly",
2041 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2042 (because of problems with the built-in one) was declared to return uschar *
2043 instead of char *, causing compiler failure.
2045 25. Fixed a file descriptor leak when processing alias/forward files.
2047 26. Fixed a minor format string issue in dbfn.c.
2049 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2051 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2052 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2055 29. When a sender address is verified, it is cached, to save repeating the test
2056 when there is more than one recipient in a message. However, when the
2057 verification involves a callout, it is possible for different callout
2058 options to be set for different recipients. It is too complicated to keep
2059 track of this in the cache, so now Exim always runs a verification when a
2060 callout is required, relying on the callout cache for the optimization.
2061 The overhead is duplication of the address routing, but this should not be
2064 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2065 to be verified with callout=postmaster, and the main callout worked but the
2066 postmaster check failed, the verification correctly failed. However, if a
2067 subsequent RCPT command asked for sender verification *without* the
2068 postmaster check, incorrect caching caused this verification also to fail,
2071 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2072 it was not caching the DNS options (qualify_single, search_parents) that
2073 were used when the lookup failed. A subsequent lookup with different
2074 options therefore always gave the same answer, though there were cases
2075 where it should not have. (Example: a "domains = !$mx_any" option on a
2076 dnslookup router: the "domains" option is always processed without any
2077 widening, but the router might have qualify_single set.) Now Exim uses the
2078 cached value only when the same options are set.
2080 32. Added John Jetmore's "exipick" utility to the distribution.
2082 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2083 than a timeout (e.g. a certificate is required, and is not provided), an
2084 Exim server now closes the connection immediately. Previously it waited for
2085 the client to close - but if the client is SSL, it seems that they each
2086 wait for each other, leading to a delay before one of them times out.
2088 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2089 maintained 0.8.x compatibility because I don't think many are using it, and
2090 it is clearly obsolete.
2092 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2095 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2096 client certificate was expired. A simple patch fixes this, though I don't
2097 understand the full logic of why the verify callback is called multiple
2100 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2101 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2102 which causes problems with some clients (such as the Certicom SSL Plus
2103 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2104 disables the coutermeasure allowing Eudora to connect."
2106 38. Exim was not checking that a write() to a log file succeeded. This could
2107 lead to Bad Things if a log got too big, in particular if it hit a file
2108 size limit. Exim now panics and dies if it cannot write to a log file, just
2109 as it does if it cannot open a log file.
2111 39. Modified OS/Makefile-Linux so that it now contains
2113 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2115 The two -D definitions ensure that Exim is compiled with large file
2116 support, which makes it possible to handle log files that are bigger than
2119 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2120 instance) a domain was checked against a named list that involved a lookup,
2121 causing $domain_data to be set, then another domain was checked against the
2122 same list, then the first domain was re-checked, the value of $domain_data
2123 after the final check could be wrong. In particular, if the second check
2124 failed, it could be set empty. This bug probably also applied to
2127 41. The strip_trailing_dot option was not being applied to the address given
2128 with the -f command-line option.
2130 42. The code for reading a message's header from the spool was incrementing
2131 $received_count, but never initializing it. This meant that the value was
2132 incorrect (doubled) while delivering a message in the same process in which
2133 it was received. In the most common configuration of Exim, this never
2134 happens - a fresh exec is done - but it can happen when
2135 deliver_drop_privilege is set.
2137 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2138 it now includes up to 150 characters of the unexpected data in the log
2141 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2142 and building data strings. The size of both of these buffers was 10 000
2143 bytes - far larger than anybody would *ever* want, thought I. Needless to
2144 say, somebody hit the limit. I have increased the maximum line length to
2145 20 000 and the maximum data length of concatenated lines to 100 000. I have
2146 also fixed two bugs, because there was no checking on these buffers. Tsk,
2147 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2148 buffer is too small.
2150 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2151 lsearch lookups. Now it does.
2153 46. When parsing a route_list item in a manualroute router, a fixed-length
2154 buffer was used for the list of hosts. I made this 1024 bytes long,
2155 thinking that nobody would ever have a list of hosts that long. Wrong.
2156 Somebody had a whole pile of complicated expansion conditions, and the
2157 string was silently truncated, leading to an expansion error. It turns out
2158 that it is easier to change to an unlimited length (owing to other changes
2159 that have happened since this code was originally written) than to build
2160 structure for giving a limitation error. The length of the item that
2161 expands into the list of hosts is now unlimited.
2163 47. The lsearch lookup could not handle data where the length of text line was
2164 more than 4095 characters. Such lines were truncated, leading to shortened
2165 data being returned. It should now handle lines of any length.
2167 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2168 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2171 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2172 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2173 no longer happens in comment lines. A list of replacements is now placed
2174 at the head of all of the source files, except those whose only change is
2175 to replace PERL_COMMAND in the very first #! line.
2177 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2178 messages on the queue, with a bottom-up merge sort, using code contributed
2179 by Michael Haardt. This should make operations like -bp somewhat faster on
2180 large queues. It won't affect queue runners, except when queue_run_in_order
2183 51. Installed eximstats 1.31 in the distribution.
2185 52. Added support for SRV lookups to the dnslookup router.
2187 53. If an ACL referred to $message_body or $message_body_end, the value was not
2188 reset for any messages that followed in the same SMTP session.
2190 54. The store-handling optimization for building very long strings was not
2191 differentiating between the different store pools. I don't think this
2192 actually made any difference in practice, but I've tidied it.
2194 55. While running the routers to verify a sender address, $sender_address
2195 was still set to the sender address. This is wrong, because when routing to
2196 send a bounce to the sender, it would be empty. Therefore, I have changed
2197 it so that, while verifying a sender address, $sender_address is set to <>.
2198 (There is no change to what happens when verifying a recipient address.)
2200 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2201 target A or AAAA records (if not already returned) without resetting the
2202 qualify_single or search_parents options of the DNS resolver. These are
2203 inappropriate in this case because the targets of MX and SRV records must
2204 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2205 target that, when qualified, matched something in the local domain. These
2206 two options are now turned off when doing these lookups.
2208 57. It seems that at least some releases of Reiserfs (which does not have the
2209 concept of a fixed number of inodes) returns zero and not -1 for the
2210 number of available inodes. This interacted badly with check_spool_inodes,
2211 which assumed that -1 was the "no such thing" setting. What I have done is
2212 to check that the total number of inodes is greater than zero before doing
2213 the test of how many are available.
2215 58. When a "warn" ACL statement has a log_message modifier, the message is
2216 remembered, and not repeated. This is to avoid a lot of repetition when a
2217 message has many recipients that cause the same warning to be written.
2218 Howewer, Exim was preserving the list of already written lines for an
2219 entire SMTP session, which doesn't seem right. The memory is now reset if a
2220 new message is started.
2222 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2223 debugging output unless log_rewrite was also set.
2225 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2226 of (int)(handle) when we know that handle contains (void *)(-1).
2228 61. The Exim daemon panic-logs an error return when it closes the incoming
2229 connection. However "connection reset by peer" seems to be common, and
2230 isn't really an error worthy of noting specially, so that particular error
2233 62. When Exim is trying to find all the local interfaces, it used to panic and
2234 die if the ioctl to get the interface flags failed. However, it seems that
2235 on at least one OS (Solaris 9) it is possible to have an interface that is
2236 included in the list of interfaces, but for which you get a failure error
2237 for this call. This happens when the interface is not "plumbed" into a
2238 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2239 failure of the "get flags" call assumes that the interface is down.
2241 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2242 makes life easier for people who are doing arithmetic on fields extracted
2243 from dates, where you often get leading zeros that should not be
2244 interpreted as octal.
2246 64. Added qualify_domain to the redirect router, to override the global
2249 65. If a pathologically long header line contained very many addresses (the
2250 report of this problem mentioned 10 000) and each of them was rewritten,
2251 Exim could use up a very large amount of memory. (It kept on making new
2252 copies of the header line as it rewrote, and never released the old ones.)
2253 At the expense of a bit more processing, the header rewriting function has
2254 been changed so that it no longer eats memory in this way.
2256 66. The generation of the Received: header has been moved from the time that a
2257 message starts to be received, to the time that it finishes. The timestamp
2258 in the Received: header should now be very close to that of the <= log
2259 line. There are two side-effects of this change:
2261 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2262 logged header lines no longer include the local Received: line, because
2263 it has not yet been created. The same applies to a copy of the message
2264 that is returned to a non-SMTP sender when a message is rejected.
2266 (b) When a filter file is tested using -bf, no additional Received: header
2267 is added to the test message. After some thought, I decided that this
2270 This change does not affect the value of $received_for. It is still set
2271 after address rewriting, but before local_scan() is called.
2273 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2275 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2276 gave an unhelpful panic error message, and a defer error. I have managed to
2277 change this behaviour so that it now rejects any supplied certificate,
2278 which seems right, as the list of acceptable certificates is empty.
2280 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2281 gave an unhelpful defer error. I have not managed to make this reject any
2282 supplied certificates, but the error message it gives is "no certificate
2283 supplied", which is not helpful.
2285 70. exigrep's output now also includes lines that are not associated with any
2286 message, but which match the given pattern. Implemented by a patch from
2287 Martin Sluka, which also tidied up the Perl a bit.
2289 71. Recipient callout verification, like sender verification, was using <> in
2290 the MAIL FROM command. This isn't really the right thing, since the actual
2291 sender may affect whether the remote host accepts the recipient or not. I
2292 have changed it to use the actual sender in the callout; this means that
2293 the cache record is now keyed on a recipient/sender pair, not just the
2294 recipient address. There doesn't seem to be a real danger of callout loops,
2295 since a callout by the remote host to check the sender would use <>.
2296 [SEE ABOVE: changed after hitting problems.]
2298 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2299 temporary errors. However, in the case of such a code being given after
2300 the end of a data transmission (i.e. after ".") Exim was failing to write
2301 a retry record for the message. (Yes, there was some broken host that was
2302 actually sending 8xx at this point.)
2304 73. An unknown lookup type in a host list could cause Exim to panic-die when
2305 the list was checked. (An example that provoked this was putting <; in the
2306 middle of a list instead of at the start.) If this happened during a DATA
2307 ACL check, a -D file could be left lying around. This kind of configuration
2308 error no longer causes Exim to die; instead it causes a defer errror. The
2309 incident is still logged to the main and panic logs.
2311 74. Buglet left over from Exim 3 conversion. The message "too many messages
2312 in one connection" was written to the rejectlog but not the mainlog, except
2313 when address rewriting (yes!) was being logged.
2315 75. Added write_rejectlog option.
2317 76. When a system filter was run not as root (that is, when system_filter_user
2318 was set), the values of the $n variables were not being returned to the
2319 main process; thus, they were not subsequently available in the $sn
2322 77. Added +return_path_on_delivery log selector.
2324 78. A connection timeout was being treated differently from recipients deferred
2325 when testing hosts_max_try with a message that was older than the host's
2326 retry timeout. (The host should not be counted, thus allowing all hosts to
2327 be tried at least once before bouncing.) This may have been the cause of an
2328 occasionally reported bug whereby a message would remain on the queue
2329 longer than the retry timeout, but would be bounced if a delivery was
2330 forced. I say "may" because I never totally pinned down the problem;
2331 setting up timeout/retry tests is difficult. See also the next item.
2333 79. The ultimate address timeout was not being applied to errors that involved
2334 a combination of host plus message (for example, a timeout on a MAIL
2335 command). When an address resolved to a number of possible hosts, and they
2336 were not all tried for each delivery (e.g. because of hosts_max_try), a
2337 message could remain on the queue longer than the retry timeout.
2339 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2342 81. Fixed an obscure SMTP outgoing bug which required at least the following
2343 conditions: (a) there was another message waiting for the same server;
2344 (b) the server returned 5xx to all RCPT commands in the first message so
2345 that the message was not completed; (c) the server dropped the connection
2346 or gave a negative response to the RSET that Exim sends to abort the
2347 transaction. The observed case was a dropped connection after DATA that had
2348 been sent in pipelining mode. That is, the server had advertised PIPELINING
2349 but was not implementing it correctly. The effect of the bug was incorrect
2350 behaviour, such as trying another host, and this could lead to a crash.
2356 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2357 and daemon.c were passed as pointers to ints; they should have been
2358 pointers to socklen_t variables (which are typically unsigned ints).
2360 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2363 3. Fixed a really odd bug that affected only the testing scheme; patching a
2364 certain fixed string in the binary changed the value of another string that
2365 happened to be identical to the end of the original first string.
2367 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2368 name", it returns that address as the IP address. On some operating
2369 systems (e.g. Solaris), it also passes back the IP address string as the
2370 "host name". However, on others (e.g. Linux), it passes back an empty
2371 string. Exim wasn't checking for this, and was changing the host name to an
2372 empty string, assuming it had been canonicized.
2374 5. Although rare, it is permitted to have more than one PTR record for a given
2375 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2376 all the names associated with an address, because they do in Solaris.
2377 However, it seems that they do not in Linux for data that comes from the
2378 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2379 I found this out when I moved to a new Linux workstation and tried to run
2380 the Exim test suite.
2382 To get round this problem I have changed the code so that it now does its
2383 own call to the DNS to look up PTR records when searching for a host name.
2384 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2385 addresses that are only in /etc/hosts are still found.
2387 This behaviour is, however, controlled by an option called host_lookup_
2388 order, which defaults to "bydns:byaddr". If people want to use the other
2389 order, or indeed, just use one or the other means of lookup, they can
2390 specify it in this variable.
2392 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2393 some operating systems, this comes back from gethostbyaddr() as an empty
2394 string, and this is what Exim used to test for. However, it seems that in
2395 other systems, "." is yielded. Exim now tests for this case too.
2397 7. The values of check_spool_space and check_log_space are now held internally
2398 as a number of kilobytes instead of an absolute number of bytes. If a
2399 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2400 kilobyte. This means that much larger values can be stored.
2402 8. Exim monitor: an attempt to get the action menu when not actually pointing
2403 at a message produces an empty menu entitled "No message selected". This
2404 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2405 no entries in it ("Shell widget menu has zero width and/or height"). So I
2406 have added a single, blank menu entry in this case.
2408 9. Added ${quote_local_part.
2410 10. MIME decoding is now applied to the contents of Subject: header lines when
2413 11. Now that a reference to $sender_host_address automatically causes a reverse
2414 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2415 host lookup before query-style lookups in lists that might use this
2416 variable. This has therefore been abolished, and the "net-" prefix is no
2417 longer necessary for query-style lookups.
2419 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2420 have been a typo for LFLAGS, so it has been changed.
2422 13. The install script calls Exim with "-C /dev/null" in order to find the
2423 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2424 to be output. Howeve, since Exim outputs its version number before the
2425 error, it didn't break the script. It just looked ugly. I fixed this by
2426 always allowing "-C /dev/null" if the caller is root.
2428 14. Ignore overlarge ACL variable number when reading spool file - insurance
2429 against a later release with more variables having written the file.
2431 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2432 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2433 this, as well as the form without the "IPv6" on the front.
2435 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2436 OS/Makefile-Darwin file.
2438 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2439 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2441 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2442 variable. However, some OS use ints for inodes. I've added cast to long int
2443 to get rid of the compiler warning.
2445 19. I had forgotten to lock out "/../" in configuration file names when
2446 ALT_CONFIG_PREFIX was set.
2448 20. Routers used for verification do not need to specify transports. However,
2449 if such a router generated a host list, and callout was configured, Exim
2450 crashed, because it could not find a port number from the (non-existent)
2451 transport. It now assumes port 25 in this circumstance.
2453 21. Added the -t option to exigrep.
2455 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2456 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2457 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2458 them is set, LOOKUP_LSEARCH is forced.
2460 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2462 24. Added sender and host information to the "rejected by local_scan()" log
2463 line; previously there was no indication of these.
2465 25. Added .include_if_exists.
2467 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2468 Linux. It turns out that not all file systems support this. Apparently some
2469 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2470 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2471 not supported on the file descriptor, is now ignored when Exim is trying to
2472 sync a directory. This applies only to Linux.
2474 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2476 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2477 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2480 29. Added support for maildirsize files from supplied patch (modified a bit).
2482 30. The use of :fail: followed by an empty string could lead Exim to respond to
2483 sender verification failures with (e.g.):
2485 550 Verification failed for <xxx>
2486 550 Sender verify failed
2488 where the first response line was missing the '-' that indicates it is not
2489 the final line of the response.
2491 31. The loop for finding the name of the user that called Exim had a hardwired
2492 limit of 10; it now uses the value of finduser_retries, which is used for
2493 all other user lookups.
2495 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2498 33. Exim was neglecting to zero errno before one call of strtol() when
2499 expanding a string and expecting an integer value. On some systems this
2500 resulted in spurious "integer overflow" errors. Also, it was casting the
2501 result into an int without checking.
2503 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2504 did not work. The code looks as if it has *never* worked, though it appears
2505 to have been documented since at least releast 1.62. I have made it work.
2507 35. The "timeout_DNS" error in retry rules, also documented since at least
2508 1.62, also never worked. As it isn't clear exactly what this means, and
2509 clearly it isn't a major issue, I have abolished the feature by treating it
2510 as "timeout", and writing a warning to the main and panic logs.
2512 36. The display of retry rules for -brt wasn't always showing the error code
2515 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2516 timeout_connect_A, timeout_connect_MX.
2518 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2519 to the empty sender.
2521 39. The daemon was not analysing the content of -oX till after it had closed
2522 stderr and disconnected from the controlling terminal. This meant that any
2523 syntax errors were only noted on the panic log, and the return code from
2524 the command was 0. By re-arranging the code a little, I've made the
2525 decoding happen first, so such errors now appear on stderr, and the return
2526 code is 1. However, the actual setting up of the sockets still happens in
2527 the disconnected process, so errors there are still only recorded on the
2530 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2531 connections (as happens on some IP stacks) was logged at start up time as
2532 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2533 it from "IPv6 and IPv4", which means that two separate sockets are being
2536 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2537 says whether AF_INET or AF_INET6 was passed as an argument.
2539 42. Exiwhat output was messed up when time zones were included in log
2542 43. Exiwhat now gives more information about the daemon's listening ports,
2543 and whether -tls-on-connect was used.
2545 44. The "port" option of the smtp transport is now expanded.
2547 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2548 silently ignored. Now an error message is written to the main and panic
2551 46. There's a new ACL modifier called "logwrite" which writes to a log file
2552 as soon as it is encountered.
2554 47. Added $local_user_uid and $local_user_gid at routing time.
2556 48. Exim crashed when trying to verify a sender address that was being
2559 49. Exim was recognizing only a space character after ".include". It now also
2560 recognizes a tab character.
2562 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2563 extracting the relevant information from the specification. The man page no
2564 longer contains scrambled data for the -d option, and I've added a section
2565 at the front about calling Exim under different names.
2567 51. Added "extra_headers" argument to the "mail" command in filter files.
2569 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2572 53. Installed eximstats 1.29.
2574 54. Added transport_filter_timeout as a generic transport option.
2576 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2577 Cc: header lines. This was required by RFC 822, but it not required by RFC
2580 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2581 incoming messages that did not have them. Now it does so only if the
2582 message originates locally, that is, if there is no associated remote host
2583 address. When Resent- header lines are present, this applies to the Resent-
2584 lines rather than the non-Resent- lines.
2586 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2587 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2589 58. Messages for configuration errors now include the name of the main
2590 configuration file - useful now that there may be more than one file in a
2591 list (.included file names were always shown).
2593 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2594 for those rare installations that do not start the daemon as root or run it
2595 setuid root. I've cut out the call to initgroups() if the daemon is not
2598 60. The Exim user and group can now be bound into the binary as text strings
2599 that are looked up at the start of Exim's processing.
2601 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2603 62. Added $mailstore_basename variable.
2605 63. Installed patch to sieve.c from Michael Haardt.
2607 64. When Exim failed to open the panic log after failing to open the main log,
2608 the original message it was trying to log was written to stderr and debug
2609 output, but if they were not available (the usual case in production), it
2610 was lost. Now it is written to syslog before the two lines that record the
2611 failures to open the logs.
2613 65. Users' Exim filters run in subprocesses under the user's uid. It is
2614 possible for a "deliver" command or an alias in a "personal" command to
2615 provoke an address rewrite. If logging of address rewriting is configured,
2616 this fails because the process is not running as root or exim. There may be
2617 a better way of dealing with this, but for the moment (because 4.30 needs
2618 to be released), I have disabled address rewrite logging when running a
2619 filter in a non-root, non-exim process.
2625 1. The buildconfig auxiliary program wasn't quoting the value set for
2626 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2627 not defined. This bug was masked in 4.22 by the effect that was fixed in
2630 2. Some messages that were rejected after a message id was allocated were
2631 shown as "incomplete" by exigrep. It no longer does this for messages that
2632 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2634 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2635 have allow_domain_literals set, the ID did not get logged in the <= line.
2636 Domain literals are now always recognized in Message-ID: header lines.
2638 4. The first argument for a ${extract expansion item is the key name or field
2639 number. Leading and trailing spaces in this item were not being ignored,
2640 causing some misleading effects.
2642 5. When deliver_drop_privilege was set, single queue runner processes started
2643 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2644 same command in the process it spins off) were not dropping privilege.
2646 6. When the daemon running as "exim" started a queue runner, it always
2647 re-executed Exim in the spun-off process. This is a waste of effort when
2648 deliver_drop_privilege is set. The new process now just calls the
2649 queue-runner function directly.
2655 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2658 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2659 ignored. Though the use of -oP was forcing the writing of a pid file, it
2660 was always written to the default place.
2662 3. If the message "no IP address found for host xxxx" is generated during
2663 incoming verification, it is now followed by identification of the incoming
2664 connection (so you can more easily find what provoked it).
2666 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2668 5. Added some features to "harden" Exim a bit more against certain attacks:
2670 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2671 be put in Local/Makefile. This is like the never_users runtime option,
2672 but it cannot be overridden. The default setting is "root".
2674 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2675 prefix string with which any file named in a -C command line option
2678 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2679 is retained for -C and -D only if the caller of Exim is root. Without
2680 it, the exim user may also use -C and -D and retain privilege.
2682 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2683 command line option is disabled.
2685 6. Macro names set by the -D option must start with an upper case letter, just
2686 like macro names defined in the configuration file.
2688 7. Added "dereference=" facility to LDAP.
2690 8. Two instances of the typo "uknown" in the source files are fixed.
2692 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2693 the Configure-Makefile script screwed up while processing it.
2695 10. Incorporated PCRE 4.4.
2697 11. The SMTP synchronization check was not operating right at the start of an
2698 SMTP session. For example, it could not catch a HELO sent before the client
2699 waited for the greeting. There is now a check for outstanding input at the
2700 point when the greeting is written. Because of the duplex, asynchronous
2701 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2702 way, but not yet received, when the check is performed.
2704 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2705 on TCP/IP sockets, because this apparently causes some broken clients to
2708 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2709 unchanged) from the Cygwin maintainer.
2711 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2712 was supported instead of testing for mbx. Effectively a typo.
2714 15. The spa authenticator server code was not checking that the input it
2715 received was valid base64.
2717 16. The debug output line for the "set" modifier in ACLs was not showing the
2718 name of the variable that was being set.
2720 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2722 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2723 lookup on its own. Something else was needed to trigger the lookup. For
2724 example, a match in host_lookup or the need for a host name in a host list.
2725 Now, if $sender_host_name is referenced and the host name has not yet been
2726 looked up, a lookup is performed. If the lookup fails, the variable remains
2727 empty, and $host_lookup_failed is set to "1".
2729 19. Added "eqi" as a case-independent comparison operator.
2731 20. The saslauthd authentication condition could segfault if neither service
2732 nor realm was specified.
2734 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2735 error message that was logged was misleading, and incoming SMTP
2736 connections were dropped. The message is now more accurate, and temporary
2737 errors are given to SMTP connections.
2739 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2740 (except RSET) with a 421 error, until QUIT is received. However, it was
2741 failing to send a response to QUIT.
2743 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2744 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2745 ACL is now run after the helo_try_verify_hosts code.
2747 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2748 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2749 case-independent, so other case variants are also recognized). Apparently
2750 some systems use these upper case variants.
2752 25. If more than two messages were waiting for the same host, and a transport
2753 filter was specified for the transport, Exim sent two messages over the
2754 same TCP/IP connection, and then failed with "socket operation on non-
2755 socket" when it tried to send the third.
2757 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2759 27. The extern definition of crypt16() in expand.c was not being excluded when
2760 the OS had its own crypt16() function.
2762 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2763 as a preferred synonym for return_size_limit, both as an option and as an
2766 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2768 30. Changed the default configuration ACL to relax the local part checking rule
2769 for addresses that are not in any local domains. For these addresses,
2770 slashes and pipe symbols are allowed within local parts, but the sequence
2771 /../ is explicitly forbidden.
2773 31. SPA server authentication was not clearing the challenge buffer before
2776 32. log_message in a "warn" ACL statement was writing to the reject log as
2777 well as to the main log, which contradicts the documentation and doesn't
2778 seem right (because no rejection is happening). So I have stopped it.
2780 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2781 However, I am unable to do any testing of this.
2783 34. Fixed an infelicity in the appendfile transport. When checking directories
2784 for a mailbox, to see if any needed to be created, it was accidentally
2785 using path names with one or more superfluous leading slashes; tracing
2786 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2788 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2789 contents are added to the log line that is written for every discarded
2790 recipient. (Previously a log_message setting was ignored.)
2792 36. The ${quote: operator now quotes the string if it is empty.
2794 37. The install script runs exim in order to find its version number. If for
2795 some reason other than non-existence or emptiness, which it checks, it
2796 could not run './exim', it was installing it with an empty version number,
2797 i.e. as "exim-". This error state is now caught, and the installation is
2800 38. An argument was missing from the function that creates an error message
2801 when Exim fails to connect to the socket for saslauthd authentication.
2802 This could cause Exim to crash, or give a corrupted message.
2804 39. Added isip, isip4, and isip6 to ${if conditions.
2806 40. The ACL variables $acl_xx are now saved with the message, and can be
2807 accessed later in routers, transports, and filters.
2809 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2810 strings in the file are not string-expanded.
2812 42. If a MAIL command specified a SIZE value that was too large to fit into an
2813 int variable, the check against message_size_limit failed. Such values are
2814 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2815 day this will have to be increased, but I don't think I want to be around
2816 when emails are that large.
2823 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2824 iconv() is not standard in FreeBSD.
2826 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2827 IPv6 enabled. The observed symptom was a segmentation fault on return from
2828 the function os_common_find_running_interfaces() in src/os.c.
2830 3. In the check_special_case() function in daemon.c I had used "errno" as an
2831 argument name, which causes warnings on some systems. This was basically a
2832 typo, since it was named "eno" in the comments!
2834 4. The code that waits for the clock to tick (at a resolution of some fraction
2835 of a second) so as to ensure message-id uniqueness was always waiting for
2836 at least one whole tick, when it could have waited for less. [This is
2837 almost certainly not relevant at current processor speeds, where it is
2838 unlikely to ever wait at all. But we try to future-proof.]
2840 5. The function that sleeps for a time interval that includes fractions of a
2841 second contained a race. It did not block SIGALRM between setting the
2842 timer, and suspending (a couple of lines later). If the interval was short
2843 and the sigsuspend() was delayed until after it had expired, the suspension
2844 never ended. On busy systems this could lead to processes getting stuck for
2847 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2848 process, before it forks any delivery processes. The open lookup caching
2849 mechanism meant that the open file or database connection was passed into
2850 the delivery process. The problem was that delivery processes always tidy
2851 up cached lookup data. This could cause a problem for the next delivery
2852 process started by the queue runner, because the external queue runner
2853 process does not know about the closure. So the next delivery process
2854 still has data in the lookup cache. In the case of a file lookup, there was
2855 no problem because closing a file descriptor in a subprocess doesn't affect
2856 the parent. However, if the lookup was caching a connection to a database,
2857 the connection was closed, and the second delivery process was likely to
2858 see errors such as "PGSQL: query failed: server closed the connection
2859 unexpectedly". The problem has been fixed by closing all cached lookups
2860 in a queue runner before running a delivery process.
2862 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2863 seem to have the "const" qualifier which it has on other OS. I've
2866 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2867 *of the same type* (client or server) with the same public name that an
2868 error should be diagnosed.
2870 9. When Exim looked up a host name for an IP address, but failed to find the
2871 original IP address when looking up the host name (a safety check), it
2872 output the message "<ip address> does not match any IP for NULL", which was
2873 confusing, to say the least. The bug was that the host name should have
2874 appeared instead of "NULL".
2876 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2877 user that is built into the binary, and the -C or -D options is used, root
2878 privilege is dropped before the configuration file is read. In addition,
2879 logging is switched to stderr instead of the normal log files. If the
2880 configuration then re-defines the Exim user, the unprivileged environment
2881 is probably not what is expected, so Exim logs a panic warning message (but
2884 However, if deliver_drop_privilege is set, the unprivileged state may well
2885 be exactly what is intended, so the warning has been cut out in that case,
2886 and Exim is allowed to try to write to its normal log files.
2892 1. smtp_return_error_details was not giving details for temporary sender
2893 or receiver verification errors.
2895 2. Diagnose a configuration error if two authenticators have the same public
2898 3. Exim used not to create the message log file for a message until the first
2899 delivery attempt. This could be confusing when incoming messages were held
2900 for policy or load reasons. The message log file is now created at the time
2901 the message is received, and an initial "Received" line is written to it.
2903 4. The automatically generated man page for command line options had a minor
2904 bug that caused no ill effects; however, a more serious problem was that
2905 the procedure for building the man page automatically didn't always
2906 operate. Consequently, release 4.20 contains an out-of-date version. This
2907 shouldn't happen again.
2909 5. When building Exim with embedded Perl support, the script that builds the
2910 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2911 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2913 6. The freeze_tell option was not being used for messages that were frozen on
2914 arrival, either by an ACL or by local_scan().
2916 7. Added the smtp_incomplete_transaction log selector.
2918 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2919 was accepting AUTH without a new EHLO.
2921 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2922 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2925 10. Logging of TCP/IP connections (when configured) now happens in the main
2926 daemon process instead of the child process, so that the TCP/IP connection
2927 count is more accurate (but it can never be perfect).
2929 11. The use of "drop" in a nested ACL was not being handled correctly in the
2930 outer ACL. Now, if condition failure induced by the nested "drop" causes
2931 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2932 or "require"), the connection is dropped.
2934 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2935 that yield "discard" can now be used with an "accept" or a "discard" verb,
2936 but an error is generated for any others (because I can't see a useful way
2937 to define what should happen).
2939 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2940 are now processed in the same way as lines in the Exim configuration file.
2941 In particular, continuation lines are supported.
2943 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2945 15. Added -ti meaning -t -i.
2947 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2948 domains, and warn by logging if others are found.
2950 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2951 returned from GIFCONF when Exim is trying to find the list of interfaces on
2952 a host. The code in os.c has been modified to copy each ifreq to an aligned
2953 structure in all cases.
2955 Also, in some cases, the returned ifreq's were being copied to a 'struct
2956 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2957 means the last couple of bytes of an IPv6 address could be chopped if the
2958 ifreq contained only a normal sockaddr (14 bytes storage).
2960 18. Named domain lists were not supported in the hosts_treat_as_local option.
2961 An entry such as +xxxx was not recognized, and was treated as a literal
2964 19. Ensure that header lines added by a DATA ACL are included in the reject log
2965 if the ACL subsequently rejects the message.
2967 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2968 MD5 (which is deprecated).
2970 21. When testing a filter file using -bf, Exim was writing a message when it
2971 took the sender from a "From " line in the message, but it was not doing so
2972 when it took $return_path from a Return-Path: header line. It now does.
2974 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2975 with a valid header line field name (a series of printing characters
2976 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2978 23. Changed "disc" in the source to "disk" to conform to the documentation and
2979 the book and for uniformity.
2981 24. Ignore Sendmail's -Ooption=value command line item.
2983 25. When execve() failed while trying to run a command in a pipe transport,
2984 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2985 could be confused with a return value of 69 from the command itself. This
2986 has been changed to 127, the value the shell returns if it is asked to run
2987 a non-existent command. The wording for the related log line suggests a
2988 non-existent command as the problem.
2990 26. If received_header_text expands to an empty string, do not add a Received:
2991 header line to the message. (Well, it adds a token one on the spool, but
2992 marks it "old" so that it doesn't get used or transmitted.)
2994 27. Installed eximstats 1.28 (addition of -nt option).
2996 28. There was no check for failure on the call to getsockname() in the daemon
2997 code. This can fail if there is a shortage of resources on the system, with
2998 ENOMEM, for example. A temporary error is now given on failure.
3000 29. Contrary to the C standard, it seems that in some environments, the
3001 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3002 program. Exim now does this explicitly; it affects the formatting of
3003 timestamps using strftime().
3005 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3006 complaints. I've now initialized all the variables, to avoid this.
3008 32. Header lines added by a system filter were not being "seen" during
3009 transport-time rewrites.
3011 33. The info_callback() function passed to OpenSSL is set up with type void
3012 (*)(SSL *, int, int), as described somewhere. However, when calling the
3013 function (actually a macro) that sets it up, the type void(*)() is
3014 expected. I've put in a cast to prevent warnings from picky compilers.
3016 34. If a DNS black list lookup found a CNAME record, but there were no A
3017 records associated with the domain it pointed at, Exim crashed.
3019 35. If a DNS black list lookup returned more than one A record, Exim ignored
3020 all but the first. It now scans all returned addresses if a particular IP
3021 value is being sought. In this situation, the contents of the
3022 $dnslist_value variable are a list of all the addresses, separated by a
3025 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3026 to accept a host name and all its aliases if the forward lookup for any of
3027 them yielded the IP address of the incoming connection. Now it accepts only
3028 those names whose forward lookup yields the correct IP address. Any other
3029 names are discarded. This closes a loophole whereby a rogue DNS
3030 administrator could create reverse DNS records to break through a
3031 wildcarded host restriction in an ACL.
3033 37. If a user filter or a system filter that ran in a subprocess used any of
3034 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3035 the wrong values were passed to the pipe command ($thisaddress had the
3036 value of $0, $0 had the value of $1, etc). This bug was introduced by
3037 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3039 38. Improved the line breaking for long SMTP error messages from ACLs.
3040 Previously, if there was no break point between 40 and 75 characters, Exim
3041 left the rest of the message alone. Two changes have been made: (a) I've
3042 reduced the minimum length to 35 characters; (b) if it can't find a break
3043 point between 35 and 75 characters, it looks ahead and uses the first one
3044 that it finds. This may give the occasional overlong line, but at least the
3045 remaining text gets split now.
3047 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3048 file descriptors might be low, and that setting 1000 would always raise it.
3049 It turns out that in some environments, the limit is already over 1000 and
3050 that lowering it causes trouble. So now Exim takes care not to decrease it.
3052 40. When delivering a message, the value of $return_path is set to $sender_
3053 address at the start of routing (routers may change the value). By an
3054 oversight, this default was not being set up when an address was tested by
3055 -bt or -bv, which affected the outcome if any router or filter referred to
3058 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3059 log only when "message" or "log_message" are set. However, if one of the
3060 conditions was an address verification, or a call to a nested ACL, the
3061 messages generated by the underlying test were being passed through. This
3062 no longer happens. The underlying message is available in $acl_verify_
3063 message for both "message" and "log_message" expansions, so it can be
3064 passed through if needed.
3066 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3067 new expansion $bh_ to give the encoded byte string without charset
3068 translation. Translation happens only if iconv() is available; HAVE_ICONV
3069 indicates this at build time. HEADERS_CHARSET gives the charset to
3070 translate to; headers_charset can change it in the configuration, and
3071 "headers charset" can change it in an individual filter file.
3073 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3074 that creates RFC 2047 encoded "words" labels them as that charset instead
3075 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3076 expansion operator; (ii) when Exim creates a From: line for a local
3077 message; (iii) when a header line is rewritten to include a "phrase" part.
3079 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3080 buggy, causing it to skip the first lines of messages whose message ID
3081 ended in 'D'. This would not have bitten before Exim release 4.14, because
3082 message IDs were unlikely to end in 'D' before then. The effect was to have
3083 incorrect size information for certain domains.
3085 45. #include "config.h" was missing at the start of the crypt16.c module. This
3086 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3089 46. If there was a timeout during a "random" callout check, Exim treated it as
3090 a failure of the random address, and carried on sending RSET and the real
3091 address. If the delay was just some slowness somewhere, the response to the
3092 original RCPT would be taken as a response to RSET and so on, causing
3093 mayhem of various kinds.
3095 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3096 when I implemented it. It didn't allow for the fact that some option values
3097 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3098 the right test for positive values.
3100 48. Domain names in DNS records are case-independent. Exim always looks them up
3101 in lower case. Some resolvers return domain names in exactly the case they
3102 appear in the zone file, that is, they may contain uppercase letters. Not
3103 all resolvers do this - some return always lower case. Exim was treating a
3104 change of case by a resolver as a change of domain, similar to a widening
3105 of a domain abbreviation. This triggered its re-routing code and so it was
3106 trying to route what was effectively the same domain again. This normally
3107 caused routing to fail (because the router wouldn't handle the domain
3108 twice). Now Exim checks for this case specially, and just changes the
3109 casing of the domain that it ultimately uses when it transmits the message
3112 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3115 50. If a filter generated a file delivery with a non-absolute name (possible if
3116 no home directory exists for the router), the forbid_file option was not
3119 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3120 the existing equality matching.
3122 52. Exim was using ints instead of ino_t variables in some places where it was
3123 dealing with inode numbers.
3125 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3126 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3127 called TMPDIR, and if it finds it is different, it changes its value.
3129 54. The smtp_printf() function is now made available to local_scan() so
3130 additional output lines can be written before returning. There is also an
3131 smtp_fflush() function to enable the detection of a dropped connection.
3132 The variables smtp_input and smtp_batched_input are exported to
3135 55. Changed the default runtime configuration: the message "Unknown user"
3136 has been removed from the ACL, and instead placed on the localuser router,
3137 using the cannot_route_message feature. This means that any verification
3138 failures that generate their own messages won't get overridden. Similarly,
3139 the "Unrouteable address" message that was in the ACL for unverifiable
3140 relay addresses has also been removed.
3142 56. Added hosts_avoid_esmtp to the smtp transport.
3144 57. The exicyclog script was not checking for the esoteric option
3145 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3146 will work only if exicyclog is run under the appropriate euid.
3148 58. Following a discussion on the list, the rules by which Exim recognises line
3149 endings on incoming messages have been changed. The -dropcr and drop_cr
3150 options are now no-ops, retained only for backwards compatibility. The
3151 following line terminators are recognized: LF CRLF CR. However, special
3152 processing applies to CR:
3154 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3155 nor a local message in the state where . is a terminator.
3157 (ii) If a bare CR is encountered in a header line, an extra space is added
3158 after the line terminator so as not to end the header. The reasoning
3159 behind this is that bare CRs in header lines are most likely either
3160 to be mistakes, or people trying to play silly games.
3162 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3163 was being incorrectly given as 18 bytes larger than it should have been.
3164 This is a VOB (very old bug).
3166 60. This may never have affected anything current, but just in case it has:
3167 When the local host is found other than at the start of a list of hosts,
3168 the local host, those with the same MX, and any that follow, are discarded.
3169 When the list in question was part of a longer list of hosts, the following
3170 hosts (not currently being processed) were also being discarded. This no
3171 longer happens. I'm not sure if this situation could ever has previously
3174 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3177 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3178 header. This is something that is recommended in a new Internet Draft, and
3179 is something that is documented as being done by Sendmail. There are two
3180 possible values. For messages generated by the autoreply transport, Exim
3183 Auto-Submitted: auto-replied
3185 whereas for all other generated messages (e.g. bounces) it adds
3187 Auto-Submitted: auto-generated
3189 63. The "personal" condition in filters now includes a test for the
3190 Auto-Submitted: header. If it contains the string "auto-" the message it
3191 not considered personal.
3193 64. Added rcpt_include_affixes as a generic transport option.
3195 65. Added queue_only_override (default true).
3197 66. Added the syslog_duplication option.
3199 67. If what should have been the first header line of a message consisted of
3200 a space followed by a colon, Exim was mis-interpreting it as a header line.
3201 It isn't of course - it is syntactically invalid and should therefore be
3202 treated as the start of the message body. The misbehaviour could have
3203 caused a number of strange effects, including loss of data in subsequent
3204 header lines, and spool format errors.
3206 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3207 client host had authenticated. This control can now be exercised by an ACL
3208 for more flexibility.
3210 69. By default, callouts do not happen when testing with -bh. There is now a
3211 variant, -bhc, which does actually run the callout code, including
3212 consulting and updating the callout cache.
3214 70. Added support for saslauthd authentication, courtesy of Alexander
3217 71. If statvfs() failed on the spool or log directories while checking their
3218 size for availability, Exim confusingly gave the error "space shortage".
3219 Furthermore, in debugging mode it crashed with a floating point exception.
3220 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3221 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3222 serious problem, Exim now writes to the main and panic logs when this
3223 happens, with details of the failure. It then refuses to accept the
3224 incoming message, giving the message "spool directory problem" or "log
3225 directory problem" with a 421 code for SMTP messages.
3227 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3228 0, 1, and 2 exist, because some OS complain for execs without them (see
3229 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3230 descriptors. However, the code omitted to check that the open succeeded,
3231 causing mysterious errors if for some reason the permissions on /dev/null
3232 got screwed. Now Exim writes a message to the main and panic logs, and
3233 bombs out if it can't open /dev/null.
3235 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3236 interact so that it is all more flexible. It is supposed to remain
3237 backwards compatible. Also added extra_local_interfaces.
3239 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3240 to bomb out with an assertion failure - to the client this appears as a
3241 connection drop. This problem occurs in the part of the code that was taken
3242 from the Samba project. Fortunately, the assertion is in a very simple
3243 function, so I have fixed this by reproducing the function inline in the
3244 one place where it is called, and arranging for authentication to fail
3245 instead of killing the process with assert().
3247 75. The SPA client code was not working when the server requested OEM rather
3248 than Unicode encoding.
3250 76. Added code to make require_files with a specific uid setting more usable in
3251 the case where statting the file as root fails - usually a non-root-mounted
3252 NFS file system. When this happens and the failure is EACCES, Exim now
3253 forks a subprocess and does the per-uid checking as the relevant uid.
3255 77. Added process_log_path.
3257 78. If log_file_path was not explicitly set, a setting of check_log_space or
3258 check_log_inodes was ignored.
3260 79. If a space check for the spool or log partitions fails, the incident is now
3261 logged. Of course, in the latter case the data may get lost...
3263 80. Added the %p formatting code to string_format() so that it can be used to
3264 print addresses in debug_print(). Adjusted all the address printing in the
3265 debugging in store.c to use %p rather than %d.
3267 81. There was a concern that a line of code in smtp_in.c could overflow a
3268 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3269 initially expressed, the concern was not well-founded, because trailing
3270 spaces are removed early. However, if the trailing spaces were followed by
3271 a NULL, they did not get removed, so the overflow was possible. Two fixes
3274 (a) I re-wrote the offending code in a cleaner fashion.
3275 (b) If an incoming SMTP command contains a NULL character, it is rejected
3278 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3279 runs initgroups(), so that if the Exim user is in any additional groups,
3280 they will be used during message reception.
3286 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob