1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.601 2010/02/23 03:01:55 jetmore Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 JJ/01 installed exipick 20100104.1, adding $max_received_linelength, $data_path, and $header_path variables; fixed documentation bugs and typos
11 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow exipick to access non-standard spools, including the "frozen" queue (Finput)
16 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body
18 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
20 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
22 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
24 NM/04 Bugzilla 915: Flag broken perl installation during build.
30 TK/01 Added patch by Johannes Berg that expands the main option
31 "spamd_address" if it starts with a dollar sign.
33 TK/02 Write list of recipients to X-Envelope-Sender header when building
34 the mbox-format spool file for content scanning (suggested by Jakob
37 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
38 (http://www.dcc-servers.net/) support via dccifd. Activated by
39 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
41 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
42 by Mark Daniel Reidel <mr@df.eu>.
44 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
45 When building exim an external PCRE library is now needed -
46 PCRE is a system library on the majority of modern systems.
47 See entry on PCRE_LIBS in EDITME file.
49 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
50 conversation. Added nologin parameter to request.
51 Patch contributed by Kirill Miazine
53 TF/01 Do not log submission mode rewrites if they do not change the address.
55 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
57 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
58 log files in place. Contributed by Roberto Lima
60 NM/04 Bugzilla 667: close socket used by dovecot authenticator
62 TF/03 Bugzilla 615: When checking the local_parts router precondition
63 after a local_part_suffix or local_part_prefix option, Exim now
64 does not use the address's named list lookup cache, since this
65 contains cached lookups for the whole local part.
67 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
68 Robert Millan. Documentation is in experimental-spec.txt
70 TF/04 Bugzilla 668: Fix parallel build (make -j).
72 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000
74 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
75 Patch provided by Jan Srzednicki
77 TF/05 Leading white space used to be stripped from $spam_report which
78 wrecked the formatting. Now it is preserved.
80 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
81 that they are available at delivery time.
83 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
85 TF/08 TLS error reporting now respects the incoming_interface and
86 incoming_port log selectors.
88 TF/09 Produce a more useful error message if an SMTP transport's hosts
89 setting expands to an empty string.
91 NM/06 Bugzilla 744: EXPN did not work under TLS.
92 Patch provided by Phil Pennock
94 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
95 Patch provided by Richard Godbee
97 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
98 acl_smtp_notquit, added index entry.
100 NM/09 Bugzilla 787: Potential buffer overflow in string_format
101 Patch provided by Eugene Bujak
103 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
104 Patch provided by Maxim Dounin
106 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
107 Patch provided by Phil Pennock
109 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
111 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
112 Patch provided by Brad "anomie" Jorsch
114 NM/14 Bugzilla 622: Added timeout setting to transport filter.
115 Patch provided by Dean Brooks
117 TK/05 Add native DKIM support (does not depend on external libraries).
119 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
120 Patch provided by Graeme Fowler
122 NM/16 Bugzilla 851: Documentation example syntax fix.
124 NM/17 Changed NOTICE file to remove references to embedded PCRE.
126 NM/18 Bugzilla 894: Fix issue with very long lines including comments in lsearch
128 NM/19 Bugzilla 745: TLS version reporting
129 Patch provided by Phil Pennock
131 NM/20 Bugzilla 167: bool: condition support
132 Patch provided by Phil Pennock
134 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken clients
135 Patch provided by Phil Pennock
137 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date
138 Patch provided by Brad "anomie" Jorsch
140 NM/23 Bugzilla 687: Fix misparses in eximstats
141 Patch provided by Heiko Schlittermann
143 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid
144 Patch provided by Heiko Schlittermann
146 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file
147 plus update to original patch.
149 NM/26 Bugzilla 799: Documentation correction for ratelimit
151 NM/27 Bugzilla 802: Improvements to local interface IP addr detection
152 Patch provided by David Brownlee
154 NM/28 Bugzilla 807: Improvements to LMTP delivery logging
156 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes
158 NM/30 Bugzilla 888: TLS documentation bugfixes
160 NM/31 Bugzilla 896: Dovecot buffer overrun fix
162 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
163 Unlike the original bugzilla I have changed all shell scripts in src tree
165 NM/33 Bugzilla 898: Transport filter timeout fix
166 Patch by Todd Rinaldo
168 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches
169 Patch by Serge Demonchaux
171 NM/35 Bugzilla 39: Base64 decode bug fixes
172 Patch by Jakob Hirsch
174 NM/36 Bugzilla 909: Correct connect() call in dcc code
176 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling
178 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed
180 NM/39 Bugzilla 911: Fixed MakeLinks build script
186 TK/01 Add preliminary DKIM support. Currently requires a forked version of
187 ALT-N's libdkim that I have put here:
188 http://duncanthrax.net/exim-experimental/
190 Note to Michael Haardt: I had to rename some vars in sieve.c. They
191 were called 'true' and it seems that C99 defines that as a reserved
192 keyword to be used with 'bool' variable types. That means you could
193 not include C99-style headers which use bools without triggering
194 build errors in sieve.c.
196 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
197 as mailq or other aliases. Changed the --help handling significantly
198 to do whats expected. exim_usage() emits usage/help information.
200 SC/01 Added the -bylocaldomain option to eximstats.
202 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
204 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
206 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
212 PH/01 Another patch from the Sieve maintainer.
214 PH/02 When an IPv6 address is converted to a string for single-key lookup
215 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
216 dots are used instead of colons so that keys in lsearch files need not
217 contain colons. This was done some time before quoting was made available
218 in lsearch files. However, iplsearch files do require colons in IPv6 keys
219 (notated using the quote facility) so as to distinguish them from IPv4
220 keys. This meant that lookups for IP addresses in host lists did not work
221 for iplsearch lookups.
223 This has been fixed by arranging for IPv6 addresses to be expressed with
224 colons if the lookup type is iplsearch. This is not incompatible, because
225 previously such lookups could never work.
227 The situation is now rather anomolous, since one *can* have colons in
228 ordinary lsearch keys. However, making the change in all cases is
229 incompatible and would probably break a number of configurations.
231 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
234 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
235 conversion specification without a maximum field width, thereby enabling
236 a rogue spamd server to cause a buffer overflow. While nobody in their
237 right mind would setup Exim to query an untrusted spamd server, an
238 attacker that gains access to a server running spamd could potentially
239 exploit this vulnerability to run arbitrary code as the Exim user.
241 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
242 $primary_hostname instead of what libspf2 thinks the hosts name is.
244 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
245 a directory entry by the name of the lookup key. Previously, if a
246 symlink pointed to a non-existing file or a file in a directory that
247 Exim lacked permissions to read, a lookup for a key matching that
248 symlink would fail. Now it is enough that a matching directory entry
249 exists, symlink or not. (Bugzilla 503.)
251 PH/03 The body_linecount and body_zerocount variables are now exported in the
254 PH/04 Added the $dnslist_matched variable.
256 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
257 This means they are set thereafter only if the connection becomes
260 PH/06 Added the client_condition to authenticators so that some can be skipped
261 by clients under certain conditions.
263 PH/07 The error message for a badly-placed control=no_multiline_responses left
264 "_responses" off the end of the name.
266 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
268 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
269 (without spaces) instead of just copying the configuration text.
271 PH/10 Added the /noupdate option to the ratelimit ACL condition.
273 PH/11 Added $max_received_linelength.
275 PH/12 Added +ignore_defer and +include_defer to host lists.
277 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
278 way in which PCRE > 7.0 is built.
280 PH/14 Implemented queue_only_load_latch.
282 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
283 MAIL command. The effect was to mangle the value on 64-bit systems.
285 PH/16 Another patch from the Sieve maintainer.
287 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
289 PH/18 If a system quota error occurred while trying to create the file for
290 a maildir delivery, the message "Mailbox is full" was not appended to the
291 bounce if the delivery eventually timed out. Change 4.67/27 below applied
292 only to a quota excession during the actual writing of the file.
294 PH/19 It seems that peer DN values may contain newlines (and other non-printing
295 characters?) which causes problems in log lines. The DN values are now
296 passed through string_printing() before being added to log lines.
298 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
299 and InterBase are left for another time.)
301 PH/21 Added message_body_newlines option.
303 PH/22 Guard against possible overflow in moan_check_errorcopy().
305 PH/23 POSIX allows open() to be a macro; guard against that.
307 PH/24 If the recipient of an error message contained an @ in the local part
308 (suitably quoted, of course), incorrect values were put in $domain and
309 $local_part during the evaluation of errors_copy.
315 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
316 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
319 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
320 issue a MAIL command.
322 PH/02 In an ACL statement such as
324 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
326 if a client was not listed at all, or was listed with a value other than
327 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
328 the condition was not true (as it should be), so access was not denied.
329 The bug was that the ! inversion was incorrectly passed on to the second
330 item. This has been fixed.
332 PH/03 Added additional dnslists conditions == and =& which are different from
333 = and & when the dns lookup returns more than one IP address.
335 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
336 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
338 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
339 FSYNC, which compiles an option called disable_fsync that allows for
340 bypassing fsync(). The documentation is heavily laced with warnings.
342 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
344 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
345 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
346 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
347 including adding "make clean"; (3) Added -fPIC when compiling the test
348 dynamically loaded module, to get rid of a warning.
350 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
351 message fails, move_frozen_messages = true and ignore_bounce_errors_after
352 = 0s. The bug is otherwise harmless.
354 PH/07 There was a bug in the dovecot authenticator such that the value of
355 $auth1 could be overwritten, and so not correctly preserved, after a
356 successful authentication. This usually meant that the value preserved by
357 the server_setid option was incorrect.
359 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
361 PH/09 Installed PCRE release 7.0.
363 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
364 run for batched SMTP input. It is now run at the start of every message
365 in the batch. While fixing this I discovered that the process information
366 (output by running exiwhat) was not always getting set for -bs and -bS
367 input. This is fixed, and it now also says "batched" for BSMTP.
369 PH/11 Added control=no_pipelining.
371 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
372 patch, slightly modified), and move the expansion of helo_data till after
373 the connection is made in the smtp transport (so it can use these
376 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
378 PH/14 Added log_selector = +pid.
380 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
382 PH/16 Add ${if forany and ${if forall.
384 PH/17 Added dsn_from option to vary the From: line in DSNs.
386 PH/18 Flush SMTP output before performing a callout, unless control =
387 no_callout_flush is set.
389 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
390 was true (the default) a successful delivery failed to delete the retry
391 item, thus causing premature timeout of the address. The bug is now
394 PH/20 Added hosts_avoid_pipelining to the smtp transport.
396 PH/21 Long custom messages for fakedefer and fakereject are now split up
397 into multiline reponses in the same way that messages for "deny" and
398 other ACL rejections are.
400 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
401 with slight modification.
403 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
404 draft, changing the syntax and factoring some duplicate code".
406 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
407 for deliveries of the second and subsequent messages over the same SMTP
410 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
411 ${reduce, with only minor "tidies".
413 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
415 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
416 expansion side effects.
418 PH/27 When a message times out after an over-quota error from an Exim-imposed
419 quota, the bounce message says "mailbox is full". This message was not
420 being given when it was a system quota that was exceeded. It now should
423 MH/03 Made $recipients available in local_scan(). local_scan() already has
424 better access to the recipient list through recipients_list[], but
425 $recipients can be useful in postmaster-provided expansion strings.
427 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
428 in the case of a MAIL command with additional options following the
429 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
430 were accidentally chopped off.
432 PH/29 SMTP synchronization checks are implemented when a command is read -
433 there is a check that no more input is waiting when there shouldn't be
434 any. However, for some commands, a delay in an ACL can mean that it is
435 some time before the response is written. In this time, more input might
436 arrive, invalidly. So now there are extra checks after an ACL has run for
437 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
438 pipelining has not been advertised.
440 PH/30 MH's patch to allow iscntrl() characters to be list separators.
442 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
443 returned in the SMTP response when smtp_return_error_details was false.
446 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
447 instead of the C I/O that was originally supplied, because problems were
450 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
451 Exim which did not show up earlier: it was assuming that a call to
452 SSL_CTX_set_info_callback() might give an error value. In fact, there is
453 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
454 was a macro that became an assignment, so it seemed to work. This has
455 changed to a proper function call with a void return, hence the compile
456 error. Exim's code has been fixed.
458 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
461 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
463 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
465 PH/37 If a message is not accepted after it has had an id assigned (e.g.
466 because it turns out to be too big or there is a timeout) there is no
467 "Completed" line in the log. When some messages of this type were
468 selected by exigrep, they were listed as "not completed". Others were
469 picked up by some special patterns. I have improved the selection
470 criteria to be more general.
472 PH/38 The host_find_failed option in the manualroute router can now be set
473 to "ignore", to completely ignore a host whose IP address cannot be
474 found. If all hosts are ignored, the behaviour is controlled by the new
475 host_all_ignored option.
477 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
478 homing or because of multiple MX records with /mx) generated more than
479 one IP address, and the following item turned out to be the local host,
480 all the secondary addresses of the first item were incorrectly removed
481 from the list, along with the local host and any following hosts (which
482 is what is supposed to happen).
484 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
485 whoever called Exim into the -H file. In the case of the daemon it was
486 behaving confusingly. When first started, it used values for whoever
487 started the daemon, but after a SIGHUP it used the Exim user (because it
488 calls itself on a restart). I have changed the code so that it now always
491 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
492 message are rejected with the same error (e.g. no authentication or bad
493 sender address), and a DATA command is nevertheless sent (as can happen
494 with PIPELINING or a stupid MUA), the error message that was given to the
495 RCPT commands is included in the rejection of the DATA command. This is
496 intended to be helpful for MUAs that show only the final error to their
499 PH/42 Another patch from the Sieve maintainer.
501 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
502 Eximstats - Fixed some broken HTML links and added missing column headers
504 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
505 columns for Rejects, Temp Rejects, Ham, and Spam rows.
507 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
509 PH/43 Yet another patch from the Sieve maintainer.
511 PH/44 I found a way to check for a TCP/IP connection going away before sending
512 the response to the final '.' that terminates a message, but only in the
513 case where the client has not sent further data following the '.'
514 (unfortunately, this is allowed). However, in many cases there won't be
515 any further data because there won't be any more messages to send. A call
516 to select() can be used: if it shows that the input is "ready", there is
517 either input waiting, or the socket has been closed. An attempt to read
518 the next input character can distinguish the two cases. Previously, Exim
519 would have sent an OK response which the client would never have see.
520 This could lead to message repetition. This fix should cure that, at
521 least in a lot of common cases.
523 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
524 advertised in response to EHLO.
530 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
531 fixed by 4.65/MH/01 (is this a record?) are fixed:
533 (i) An empty string was always treated as zero by the numeric comparison
534 operators. This behaviour has been restored.
536 (ii) It is documented that the numeric comparison operators always treat
537 their arguments as decimal numbers. This was broken in that numbers
538 starting with 0 were being interpreted as octal.
540 While fixing these problems I realized that there was another issue that
541 hadn't been noticed. Values of message_size_limit (both the global option
542 and the transport option) were treated as octal if they started with 0.
543 The documentation was vague. These values are now always treated as
544 decimal, and I will make that clear in the documentation.
550 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
551 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
554 MH/01 Don't check that the operands of numeric comparison operators are
555 integers when their expansion is in "skipping" mode (fixes bug
556 introduced by 4.64-PH/07).
558 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
559 child addresses, Exim now panics and dies. Previously, because the count
560 is held in a short int, deliveries were likely to be lost. As such a
561 large number of recipients for a single message is ridiculous
562 (performance will be very, very poor), I have chosen to impose a limit
563 rather than extend the field.
569 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
570 leftover -K file (the existence of which was triggered by #402).
571 While we were at it, introduced process PID as part of the -K
572 filename. This should rule out race conditions when creating
575 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
576 processing considerably. Previous code took too long for large mails,
577 triggering a timeout which in turn triggers #401.
579 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
580 in the DK code in transports.c. sendfile() is not really portable,
581 hence the _LINUX specificness.
583 TF/01 In the add_headers option to the mail command in an Exim filter,
584 there was a bug that Exim would claim a syntax error in any
585 header after the first one which had an odd number of characters
588 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
589 callout verification, Exim cached a "reject" for the entire domain. This
590 is correct for most verifications, but it is not correct for a recipient
591 verification with use_sender or use_postmaster set, because in that case
592 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
593 case of MAIL FROM:<> rejection from other early rejections (e.g.
594 rejection of HELO). When verifying a recipient using a non-null MAIL
595 address, the cache is ignored if it shows MAIL FROM:<> rejection.
596 Whatever the result of the callout, the value of the domain cache is
597 left unchanged (for any other kind of callout, getting as far as trying
598 RCPT means that the domain itself is ok).
600 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
603 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
604 manifest itself as EPIPE rather than ECONNECT. When tidying away a
605 session, the daemon ignores ECONNECT errors and logs others; it now
606 ignores EPIPE as well.
608 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
609 (quoted-printable decoding).
611 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
612 later the small subsequent patch to fix an introduced bug.
614 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
616 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
618 PH/08 An error is now given if message_size_limit is specified negative.
620 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
621 to be given (somewhat) arbitrary names.
623 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
626 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
627 miscellaneous code fixes
629 PH/10 Added the log_reject_target ACL modifier to specify where to log
632 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
633 hostname. This is wrong, because it relates to the incoming message (and
634 probably the interface on which it is arriving) and not to the outgoing
635 callout (which could be using a different interface). This has been
636 changed to use the value of the helo_data option from the smtp transport
637 instead - this is what is used when a message is actually being sent. If
638 there is no remote transport (possible with a router that sets up host
639 addresses), $smtp_active_hostname is used.
641 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
642 tweaks were necessary in order to get it to work (see also 21 below):
643 (a) The code assumed that strncpy() returns a negative number on buffer
644 overflow, which isn't the case. Replaced with Exim's string_format()
646 (b) There were several signed/unsigned issues. I just did the minimum
647 hacking in of casts. There is scope for a larger refactoring.
648 (c) The code used strcasecmp() which is not a standard C function.
649 Replaced with Exim's strcmpic() function.
650 (d) The code set only $1; it now sets $auth1 as well.
651 (e) A simple test gave the error "authentication client didn't specify
652 service in request". It would seem that Dovecot has changed its
653 interface. Fortunately there's a specification; I followed it and
654 changed what the client sends and it appears to be working now.
656 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
659 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
660 address A is aliased to B and C, where B exists and C does not. Without
661 -v the output is "A verified" because verification stops after a
662 successful redirection if more than one address is generated. However,
663 with -v the child addresses are also verified. Exim was outputting "A
664 failed to verify" and then showing the successful verification for C,
665 with its parentage. It now outputs "B failed to verify", showing B's
666 parentage before showing the successful verification of C.
668 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
669 look up a TXT record in a specific list after matching in a combined
672 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
673 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
674 they consult the DNS. I had assumed they would set it the way they
675 wanted; and indeed my experiments on Linux seem to show that in some
676 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
677 To be on the safe side, however, I have now made the interface to
678 host_find_byname() similar to host_find_bydns(), with an argument
679 containing the DNS resolver options. The host_find_byname() function now
680 sets these options at its start, just as host_find_bydns() does. The smtp
681 transport options dns_qualify_single and dns_search_parents are passed to
682 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
683 of host_find_byname() use the default settings of RES_DEFNAMES
684 (qualify_single) but not RES_DNSRCH (search_parents).
686 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
687 spool_read_header() do less string testing, by means of a preliminary
688 switch on the second character of optional "-foo" lines. (This is
689 overdue, caused by the large number of possibilities that now exist.
690 Originally there were few.) While I was there, I also converted the
691 str(n)cmp tests so they don't re-test the leading "-" and the first
692 character, in the hope this might squeeze out yet more improvement.
694 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
695 flag allowing group syntax was set by the header_syntax check but not
696 turned off, possible causing trouble later; (2) The flag was not being
697 set at all for the header_verify test, causing "group"-style headers to
698 be rejected. I have now set it in this case, and also caused header_
699 verify to ignore an empty address taken from a group. While doing this, I
700 came across some other cases where the code for allowing group syntax
701 while scanning a header line wasn't quite right (mostly, not resetting
702 the flag correctly in the right place). These bugs could have caused
703 trouble for malformed header lines. I hope it is now all correct.
705 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
706 with the "reply" argument non-NULL. The code, however (which originally
707 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
708 but it didn't always do it. This confused somebody who was copying the
709 code for some other use. I have removed all the tests.
711 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
712 feature that was used to support insecure browsers during the U.S. crypto
713 embargo. It requires special client support, and Exim is probably the
714 only MTA that supported it -- and would never use it because real RSA is
715 always available. This code has been removed, because it had the bad
716 effect of slowing Exim down by computing (never used) parameters for the
717 RSA_EXPORT functionality.
719 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
720 authenticator to fail if there's a tab character in the incoming data
721 (there should never be unless someone is messing about, as it's supposed
722 to be base64-encoded). Also added, on Timo's advice, the "secured" option
723 if the connection is using TLS or if the remote IP is the same as the
724 local IP, and the "valid-client-cert option" if a client certificate has
727 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
728 authenticators. This can be used for authorization after authentication
729 succeeds. (In the case of plaintext, it servers for both authentication
732 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
733 if any retry times were supplied.
735 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
736 connection, where there is no client IP address to check. In this
737 situation, the verify now always succeeds.
739 PH/25 Applied John Jetmore's -Mset patch.
741 PH/26 Added -bem to be like -Mset, but loading a message from a file.
743 PH/27 In a string expansion for a processed (not raw) header when multiple
744 headers of the same name were present, leading whitespace was being
745 removed from all of them, but trailing whitespace was being removed only
746 from the last one. Now trailing whitespace is removed from each header
747 before concatenation. Completely empty headers in a concatenation (as
750 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
751 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
753 PH/29 [Removed. This was a change that I later backed out, and forgot to
754 correct the ChangeLog entry (that I had efficiently created) before
755 committing the later change.]
757 PH/30 Exim was sometimes attempting to deliver messages that had suffered
758 address errors (4xx response to RCPT) over the same connection as other
759 messages routed to the same hosts. Such deliveries are always "forced",
760 so retry times are not inspected. This resulted in far too many retries
761 for the affected addresses. The effect occurred only when there were more
762 hosts than the hosts_max_try setting in the smtp transport when it had
763 the 4xx errors. Those hosts that it had tried were not added to the list
764 of hosts for which the message was waiting, so if all were tried, there
765 was no problem. Two fixes have been applied:
767 (i) If there are any address or message errors in an SMTP delivery, none
768 of the hosts (tried or untried) are now added to the list of hosts
769 for which the message is waiting, so the message should not be a
770 candidate for sending over the same connection that was used for a
771 successful delivery of some other message. This seems entirely
772 reasonable: after all the message is NOT "waiting for some host".
773 This is so "obvious" that I'm not sure why it wasn't done
774 previously. Hope I haven't missed anything, but it can't do any
775 harm, as the worst effect is to miss an optimization.
777 (ii) If, despite (i), such a delivery is accidentally attempted, the
778 routing retry time is respected, so at least it doesn't keep
779 hammering the server.
781 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
782 in ${readsocket because some servers need this prod.
784 PH/32 Added some extra debug output when updating a wait-xxx database.
786 PH/33 The hint "could be header name not terminated by colon", which has been
787 given for certain expansion errors for a long time, was not being given
788 for the ${if def:h_colon_omitted{... case.
790 PH/34 The spec says: "With one important exception, whenever a domain list is
791 being scanned, $domain contains the subject domain." There was at least
792 one case where this was not true.
794 PH/35 The error "getsockname() failed: connection reset by peer" was being
795 written to the panic log as well as the main log, but it isn't really
796 panic-worthy as it just means the connection died rather early on. I have
797 removed the panic log writing for the ECONNRESET error when getsockname()
800 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
801 runs only) independently of the message's sender address. This meant
802 that, if the 4xx error was in fact related to the sender, a different
803 message to the same recipient with a different sender could confuse
804 things. In particualar, this can happen when sending to a greylisting
805 server, but other circumstances could also provoke similar problems.
806 I have changed the default so that the retry time for these errors is now
807 based a combination of the sender and recipient addresses. This change
808 can be overridden by setting address_retry_include_sender=false in the
811 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
812 remote server are returned as part of bounce messages. This was not
813 happening for LMTP over a pipe (the lmtp transport), but now it is the
814 same for both kinds of LMTP.
816 PH/38 Despite being documented as not happening, Exim was rewriting addresses
817 in header lines that were in fact CNAMEs. This is no longer the case.
819 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
820 and queue runs started by the daemon processed all messages. This has
821 been fixed so that -R and -S can now usefully be given with -q<time>.
823 PH/40 Import PCRE release 6.7 (fixes some bugs).
825 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
827 PH/42 Give an error if -q is specified more than once.
829 PH/43 Renamed the variables $interface_address and $interface_port as
830 $received_ip_address and $received_port, to make it clear that these
831 values apply to message reception, and not to the outgoing interface when
832 a message is delivered. (The old names remain recognized, of course.)
834 PH/44 There was no timeout on the connect() call when using a Unix domain
835 socket in the ${readsocket expansion. There now is.
837 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
838 be meaningful with "accept".
840 SC/01 Eximstats V1.43
841 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
843 SC/02 Eximstats V1.44
844 Use a glob alias rather than an array ref in the generated
845 parser. This improves both readability and performance.
847 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
848 Collect SpamAssassin and rejection statistics.
849 Don't display local sender or destination tables unless
850 there is data to show.
851 Added average volumes into the top table text output.
853 SC/04 Eximstats V1.46
854 Collect data on the number of addresses (recipients)
855 as well as the number of messages.
857 SC/05 Eximstats V1.47
858 Added 'Message too big' to the list of mail rejection
859 reasons (thanks to Marco Gaiarin).
861 SC/06 Eximstats V1.48
862 Mainlog lines which have GMT offsets and are too short to
863 have a flag are now skipped.
865 SC/07 Eximstats V1.49 (Alain Williams)
866 Added the -emptyok flag.
868 SC/08 Eximstats V1.50
869 Fixes for obtaining the IP address from reject messages.
871 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
872 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
873 whitesspace changes from 4.64-PH/27
875 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
878 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
879 are found, allow negative numbers in numeric criteria)
881 JJ/06 exipick.20061117.2, added new $message_body_missing variable
883 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
884 to match changes made in 4.64-PH/43
886 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
888 PH/47 Put in an explicit test for a DNS lookup of an address record where the
889 "domain" is actually an IP address, and force a failure. This locks out
890 those revolvers/nameservers that support "A-for-A" lookups, in
891 contravention of the specifications.
893 PH/48 When a host name was looked up from an IP address, and the subsequent
894 forward lookup of the name timed out, the host name was left in
895 $sender_host_name, contrary to the specification.
897 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
898 restricted to single-key lookups, Exim was not diagnosing an error if
899 * or *@ was used with a query-style lookup.
901 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
903 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
904 long ago, but noone interested enough thought of it. Let's just say that
905 the "1.1" means that there are some new functions that weren't there at
906 some point in the past.
908 PH/51 Error processing for expansion failure of helo_data from an smtp
909 transport during callout processing was broken.
911 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
912 tested/used via the -bh/-bhc/-bs options.
914 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
915 bug, fixed in subsequent PCRE releases).
917 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
918 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
920 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
926 SC/01 Use a glob alias rather than an array ref in eximstats generated
927 parser. This improves both readability and performance.
929 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
930 Don't display local sender or destination tables in eximstats unless
931 there is data to show.
932 Added average volumes into the eximstats top table text output.
934 SC/03 Collect data on the number of addresses (recipients) as well
935 as the number of messages in eximstats.
937 TF/01 Correct an error in the documentation for the redirect router. Exim
938 does (usually) call initgroups() when daemonizing.
940 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
941 with consistent privilege compared to when running as a daemon.
943 TF/03 Note in the spec that $authenticated_id is not set for local
944 submissions from trusted users.
946 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
947 Thanks to Dean Brooks <dean@iglou.com> for the patch.
949 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
950 by adding some example configuration directives to the default
951 configuration file. A little bit of work is required to uncomment the
952 directives and define how usernames and passwords are checked, but
953 there is now a framework to start from.
955 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
956 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
957 without this. I don't know how relevant this is to other LDAP libraries.
959 PH/02 Add the verb name to the "unknown ACL verb" error.
961 PH/03 Magnus Holmgren's patch for filter_prepend_home.
963 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
965 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
966 directory not expanded when it should be if an expanded home directory
967 was set for the address (which is overridden by the transport).
969 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
972 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
973 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
974 because it is too late at that time, and has no effect.
976 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
977 security issue with \' (bugzilla #107). I could not use the
978 PQescapeStringConn() function, because it needs a PGconn value as one of
981 PH/08 When testing addresses using -bt, indicate those final addresses that
982 are duplicates that would not cause an additional delivery. At least one
983 person was confused, thinking that -bt output corresponded to deliveries.
984 (Suppressing duplicates isn't a good idea as you lose the information
985 about possibly different redirections that led to the duplicates.)
987 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
988 systems where poll() doesn't work, in particular OS X.
990 PH/10 Added more information to debugging output for retry time not reached.
992 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
993 operations in malware.c.
995 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
998 PH/13 If write_rejectlog was set false when logging was sent to syslog with
999 syslog_duplication set false, log lines that would normally be written
1000 both the the main log and to the reject log were not written to syslog at
1003 PH/14 In the default configuration, change the use of "message" in ACL warn
1004 statements to "add_header".
1006 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1007 not followed by a command (e.g. "seen endif").
1009 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1010 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1013 PH/17 Added extra conditions to the default value of delay_warning_condition
1017 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1018 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1019 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1022 The Auto-Submitted: and various List- headers are standardised, whereas I
1023 don't think Precedence: ever was.
1025 PH/18 Refactored debugging code in route_finduser() to show more information,
1026 in particular, the error code if getpwnam() issues one.
1028 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1029 This is apparently needed in addition to the PH/07 change above to avoid
1030 any possible encoding problems.
1032 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1033 but not after initializing Perl.
1035 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1036 output them only if debugging. By default they are written stderr,
1037 apparently, which is not desirable.
1039 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1042 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1045 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1047 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1048 authenticated or an ident call has been made. Suppress the default
1049 values for $authenticated_id and $authenticated_sender (but permit -oMai
1050 and -oMas) when testing with -bh.
1052 PH/24 Re-jigged the order of the tests in the default configuration so that the
1053 tests for valid domains and recipients precede the DNS black list and CSA
1054 tests, on the grounds that those ones are more expensive.
1056 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1057 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1058 This bug exists in every version of Exim that I still have, right back to
1061 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1062 However, an attempt to turn on case-sensitivity in a regex key by
1063 including (?-i) didn't work because the subject string was already
1064 lowercased, and the effects were non-intuitive. It turns out that a
1065 one-line patch can be used to allow (?-i) to work as expected.
1071 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1072 other effects) broke the use of negated acl sub-conditions.
1074 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1077 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1078 "Deny" causes Exim to reject the incoming connection with a 554 error.
1079 Unfortunately, if there is a major crisis, such as a disk failure,
1080 tcp-wrappers gives "deny", whereas what one would like would be some
1081 kind of temporary error. A kludge has been added to help with this.
1082 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1083 554 error is used if errno is still zero or contains ENOENT (which occurs
1084 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1087 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1089 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1090 errors. Otherwise a message that provokes a temporary error (when other
1091 messages do not) can cause a whole host to time out.
1093 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1094 addresses were routed directly to files or pipes from a redirect router.
1095 File deliveries just didn't batch; pipe deliveries might have suffered
1098 PH/06 A failure to get a lock for a hints database would erroneously always say
1099 "Failed to get write lock", even when it was really a read lock.
1101 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1102 of 0600. This has been changed to use the value of the lockfile_mode
1103 option (which defaults to 0600).
1105 PH/08 Applied small patch from the Sieve maintainer.
1107 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1108 folder from quota calculations, a direct delivery into this folder messed
1109 up the contents of the maildirsize file. This was because the regex was
1110 used only to exclude .Trash (or whatever) when the size of the mailbox
1111 was calculated. There was no check that a delivery was happening into an
1112 excluded directory. This bug has been fixed by ignoring all quota
1113 processing for deliveries into excluded directories.
1115 PH/10 Added the maildirfolder_create_regex option to appendfile.
1121 PH/01 The code for finding all the local interface addresses on a FreeBSD
1122 system running IPv6 was broken. This may well have applied to all BSD
1123 systems, as well as to others that have similar system calls. The broken
1124 code found IPv4 interfaces correctly, but gave incorrect values for the
1125 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1126 that it would not match correctly against @[] and not recognize the IPv6
1129 PH/02 The ipliteral router was not recognizing addresses of the form user@
1130 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1132 PH/03 Added disable_ipv6.
1134 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1135 decoded form, because it is most often used to construct To: headers
1136 lines in autoreplies, and the decoded form may well be syntactically
1137 invalid. However, $reply_address has leading white space removed, and all
1138 newlines turned into spaces so that the autoreply transport does not
1141 PH/05 If group was specified without a user on a router, and no group or user
1142 was specified on a transport, the group from the router was ignored.
1144 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1145 for visible compile-time settings that can be used to change these
1146 numbers, for those that want even more. Backwards compatibility with old
1147 spool files has been maintained. However, going back to a previous Exim
1148 release will lost any variables that are in spool files.
1150 PH/07 Two small changes when running in the test harness: increase delay when
1151 passing a TCP/IP connection to a new process, in case the original
1152 process has to generate a bounce, and remove special handling of
1153 127.0.0.2 (sic), which is no longer necessary.
1155 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1156 be the same on different OS.
1158 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1161 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1162 whether --show-vars was specified or not
1164 JJ/02 exipick: Added support for new ACL variable spool format introduced
1167 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1168 syntactically invalid From: or Reply-to: line, and a filter used this to
1169 generate an autoreply, and therefore failed to obtain an address for the
1170 autoreply, Exim could try to deliver to a non-existent relative file
1171 name, causing unrelated and misleading errors. What now happens is that
1172 it logs this as a hard delivery error, but does not attempt to create a
1175 PH/11 The exinext utility has a -C option for testing purposes, but although
1176 the given file was scanned by exinext itself; it wasn't being passed on
1177 when Exim was called.
1179 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1180 an end-of-file indication when reading a command response.
1182 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1183 compiled. In many other places in Exim, IPv6 addresses are always
1184 recognized, so I have changed this. It also means that IPv4 domain
1185 literals of the form [IPV4:n.n.n.n] are now always recognized.
1187 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1188 used if the router is not running as root, for example, when verifying at
1189 ACL time, or when using -bh. The debugging output from this situation was
1190 non-existent - all you got was a failure to exec. I have made two
1193 (a) Failures to set uid/gid, the current directory, or a process leader
1194 in a subprocess such as that created by queryprogram now generate
1195 suitable debugging ouput when -d is set.
1197 (b) The queryprogram router detects when it is not running as root,
1198 outputs suitable debugging information if -d is set, and then runs
1199 the subprocess without attempting to change uid/gid.
1201 PH/15 Minor change to Makefile for building test_host (undocumented testing
1204 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1205 additional section of a DNS packet that returns MX or SRV records.
1206 Instead, it always explicitly searches for A/AAAA records. This avoids
1207 major problems that occur when a DNS server includes only records of one
1208 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1209 fixed another bug: if SRV records were looked up and the corresponding
1210 address records were *not* found in the additional section, the port
1211 values from the SRV records were lost.
1213 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1214 using the correct key (the original address) when searching the retry
1215 rules in order to find which one to use for generating the retry hint.
1217 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1218 adding the default one. Similarly, if it contains a Reply-To: header, the
1219 errors_reply_to option, if set, is not used.
1221 PH/19 When calculating a retry time, Exim used to measure the "time since
1222 failure" by looking at the "first failed" field in the retry record. Now
1223 it does not use this if it is later than than the arrival time of the
1224 message. Instead it uses the arrival time. This makes for better
1225 behaviour in cases where some deliveries succeed, thus re-setting the
1226 "first failed" field. An example is a quota failure for a huge message
1227 when small messages continue to be delivered. Without this change, the
1228 "time since failure" will always be short, possible causing more frequent
1229 delivery attempts for the huge message than are intended.
1230 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1232 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1233 $1, $2, $3) because the numerical variables can be reset during some
1234 expansion items (e.g. "match"), thereby losing the authentication data.
1236 PH/21 Make -bV show the size of off_t variables so that the test suite can
1237 decide whether to run tests for quotas > 2G.
1239 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1240 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1241 filecount value is greater than 2G or if a quota value is greater than 2G
1242 on a system where the size of off_t is not greater than 4, a panic error
1245 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1246 never match. The debug and -bh output now contains an explicit error
1247 message indicating a malformed IPv4 address or mask.
1249 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1250 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1251 PH/23 above applies.
1253 PH/25 Do not write to syslog when running in the test harness. The only
1254 occasion when this arises is a failure to open the main or panic logs
1255 (for which there is an explicit test).
1257 PH/26 Added the /no_tell option to "control=freeze".
1259 PH/27 If a host name lookup failed very early in a connection, for example, if
1260 the IP address matched host_lookup and the reverse lookup yielded a name
1261 that did not have a forward lookup, an error message of the form "no IP
1262 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1263 could be logged. Now it outputs the IP address instead of "NULL".
1265 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1266 allows the sender and the authenticated sender to be set when
1267 submitting a message from within Exim. Since child_open_exim() is
1268 documented for local_scan(), the new function should be too.
1270 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1271 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1272 results in an empty string is now treated as unset.
1274 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1276 PH/31 Added sender_verify_fail logging option.
1278 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1279 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1280 tidied the source and removed it altogether.
1282 PH/33 When a queue run was abandoned because the load average was too high, a
1283 log line was always written; now it is written only if the queue_run log
1284 selector is set. In addition, the log line for abandonment now contains
1285 information about the queue run such as the pid. This is always present
1286 in "start" and "stop" lines but was omitted from the "abandon" line.
1288 PH/34 Omit spaces between a header name and the colon in the error message that
1289 is given when verify = headers_syntax fails (if there are lots of them,
1290 the message gets confusing).
1292 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1293 names, as there are now some PTR records that contain slashes. This check
1294 is only to protect against broken name servers that fall over on strange
1295 characters, so the fact that it applies to all lookups doesn't matter.
1297 PH/36 Now that the new test suite is complete, we can remove some of the
1298 special code in Exim that was needed for the old test suite. For example,
1299 sorting DNS records because real resolvers return them in an arbitrary
1300 order. The new test suite's fake resolver always returns records in the
1303 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1304 bounces) except when queue_only is set, to avoid logging races between
1305 the different processes.
1307 PH/38 Panic-die if .include specifies a non-absolute path.
1309 PH/39 A tweak to the "H" retry rule from its user.
1311 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1312 a label. They prevented compilation on older perls.
1314 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1315 a warning to be raised on newish perls.
1317 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1318 on queue. Changes to match documented behaviour of showing count of
1319 messages matching specified criteria.
1321 PH/40 Changed the default ident timeout from 30s to 5s.
1323 PH/41 Added support for the use of login_cap features, on those BSD systems
1324 that have them, for controlling the resources used by pipe deliveries.
1326 PH/42 The content-scanning code uses fopen() to create files in which to put
1327 message data. Previously it was not paying any attention to the mode of
1328 the files. Exim runs with umask(0) because the rest of the code creates
1329 files with open(), and sets the required mode explicitly. Thus, these
1330 files were ending up world-writeable. This was not a big issue, because,
1331 being within the spool directory, they were not world-accessible. I have
1332 created a function called modefopen, which takes an additional mode
1333 argument. It sets umask(777), creates the file, chmods it to the required
1334 mode, then resets the umask. All the relevant calls to fopen() in the
1335 content scanning code have been changed to use this function.
1337 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1338 to 24 hours. This avoids potential overflow problems when processing G
1339 and H retry rules. I suspect nobody ever tinkers with this value.
1341 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1343 PH/45 When the plaintext authenticator is running as a client, the server's
1344 challenges are checked to ensure they are valid base64 strings. By
1345 default, the authentication attempt is cancelled if an invalid string is
1346 received. Setting client_ignore_invalid_base64 true ignores these errors.
1347 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1348 they are received. Thus, the responses can be made to depend on the
1349 challenges. If an invalid string is ignored, an empty string is placed in
1352 PH/46 Messages that are created by the autoreply transport now contains a
1353 References: header, in accordance with RFCs 2822 and 3834.
1355 PH/47 Added authenticated_sender_force to the smtp transport.
1357 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1359 PH/49 Installed latest patch from the Sieve maintainer.
1361 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1362 was also set, the appendfile transport was unnecessarily scanning a
1363 directory of message files (e.g. for maildir delivery) to find the count
1364 of files (along with the size), even though it did not need this
1365 information. It now does the scan only if it needs to find either the
1366 size of the count of files.
1368 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1370 PH/52 Two bugs concerned with error handling when the smtp transport is
1373 (i) Exim was not creating retry information for temporary errors given
1374 for individual recipients after the DATA command when the smtp transport
1375 was used in LMTP mode. This meant that they could be retried too
1376 frequently, and not timed out correctly.
1378 (ii) Exim was setting the flag that allows error details to be returned
1379 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1380 recipients that were returned after the DATA command.
1382 PH/53 This is related to PH/52, but is more general: for any failing address,
1383 when detailed error information was permitted to be returned to the
1384 sender, but the error was temporary, then after the final timeout, only
1385 "retry timeout exceeded" was returned. Now it returns the full error as
1386 well as "retry timeout exceeded".
1388 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1389 do this, and (what is worse) MTAs that accept it.
1391 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1392 will now be deprecated.
1394 PH/56 New os.c-cygwin from the Cygwin maintainer.
1396 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1397 formats (previously only available in exim formats via -bpr, -bpru,
1398 and -bpra. Now also available in native and exiqgrep formats)
1400 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1401 with very large, slow to parse queues
1403 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1405 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1407 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1408 responses to authentication challenges, though it was showing the
1409 challenges; (ii) I've removed the CR characters from the debug output for
1412 PH/58 Allow for the insertion of a newline as well as a space when a string
1413 is turned into more than one encoded-word during RFC 2047 encoding. The
1414 Sieve code now uses this.
1416 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1417 data_4xx, lost_connection, tls_required.
1419 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1420 message was being sent as an SMTP response.
1422 PH/61 Add -l and -k options to exicyclog.
1424 PH/62 When verifying, if an address was redirected to one new address, so that
1425 verification continued, and the new address failed or deferred after
1426 having set something in $address_data, the value of $address_data was not
1427 passed back to the ACL. This was different to the case when no
1428 redirection occurred. The value is now passed back in both cases.
1430 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1431 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1432 use login_cap.h, so on its own it isn't the distinguishing feature. The
1433 new name refers directly to the setclassresources() function.
1435 PH/65 Added configuration files for NetBSD3.
1437 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1439 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1440 is preferred over IPv4.
1442 PH/68 The bounce_return_message and bounce_return_body options were not being
1443 honoured for bounces generated during the reception of non-SMTP messages.
1444 In particular, this applied to messages rejected by the ACL. This bug has
1445 been fixed. However, if bounce_return_message is true and bounce_return_
1446 body is false, the headers that are returned for a non-SMTP message
1447 include only those that have been read before the error was detected.
1448 (In the case of an ACL rejection, they have all been read.)
1450 PH/69 The HTML version of the specification is now built in a directory called
1451 spec_html instead of spec.html, because the latter looks like a path with
1452 a MIME-type, and this confuses some software.
1454 PH/70 Catch two compiler warnings in sieve.c.
1456 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1457 function verify_get_ident() calls ip_connect() to connect a socket, but
1458 if the "connect()" function timed out, ip_connect() used to close the
1459 socket. However, verify_get_ident() also closes the socket later, and in
1460 between Exim writes to the log, which may get opened at this point. When
1461 the socket was closed in ip_connect(), the log could get the same file
1462 descriptor number as the socket. This naturally causes chaos. The fix is
1463 not to close the socket in ip_connect(); the socket should be closed by
1464 the function that creates it. There was only one place in the code where
1465 this was missing, in the iplookup router, which I don't think anybody now
1466 uses, but I've fixed it anyway.
1468 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1469 well as to direct DNS lookups. Otherwise the handling of names in host
1470 lists is inconsistent and therefore confusing.
1476 PH/01 Two changes to the default runtime configuration:
1478 (1) Move the checks for relay_from_hosts and authenticated clients from
1479 after to before the (commented out) DNS black list checks.
1481 (2) Add control=submission to the relay_from_hosts and authenticated
1482 clients checks, on the grounds that messages accepted by these
1483 statements are most likely to be submissions.
1485 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1487 (1) Generate an error if the third argument for the ${prvs expansion is
1490 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1493 (3) Reset the variables that are obtained from the first argument of
1494 ${prvscheck and used in the second argument before leaving the code,
1495 because their memory is reclaimed, so using them afterwards may do
1498 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1499 one (it's much easier than Tom thought :-).
1501 (5) Because of (4), we can now allow for the use of $prvscheck_result
1502 inside the third argument.
1504 PH/03 For some reason, the default setting of PATH when running a command from
1505 a pipe transport was just "/usr/bin". I have changed it to
1508 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1509 anything to be listed in the output from -bV.
1511 PH/05 When a filter generated an autoreply, the entire To: header line was
1512 quoted in the delivery log line, like this:
1514 => >A.N.Other <ano@some.domain> <original@ddress> ...
1516 This has been changed so that it extracts the operative address. There
1517 may be more than one such address. If so, they are comma-separated, like
1520 => >ano@some.domain,ona@other.domain <original@ddress> ...
1522 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1523 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1524 not being looked up in the rDNS to get a host name, Exim was showing the
1525 IP address twice in Received: lines, even though the IP addresses were
1526 identical. For example:
1528 Received: from [1.2.3.4] (helo=[1.2.3.4])
1530 However, if the real host name was known, it was omitting the HELO data
1531 if it matched the actual IP address. This has been tidied up so that it
1532 doesn't show the same IP address twice.
1534 PH/07 When both +timestamp and +memory debugging was on, the value given by
1535 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1536 called by the debug printing, thereby overwriting the timestamp buffer.
1537 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1539 PH/08 When the original message was included in an autoreply transport, it
1540 always said "this is a copy of the message, including all the headers",
1541 even if body_only or headers_only was set. It now gives an appropriate
1544 PH/09 Applied a patch from the Sieve maintainer which:
1546 o fixes some comments
1547 o adds the (disabled) notify extension core
1548 o adds some debug output for the result of if/elsif tests
1549 o points to the current vacation draft in the documentation
1550 and documents the missing references header update
1554 o fixes a bug in processing the envelope test (when testing
1555 multiple envelope elements, the last element determinted the
1558 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1559 Electronic Mail") by including:
1561 Auto-submitted: auto-generated
1563 in the messages that it generates (bounce messages and others, such as
1564 warnings). In the case of bounce messages for non-SMTP mesages, there was
1565 also a typo: it was using "Auto_submitted" (underscore instead of
1566 hyphen). Since every message generated by Exim is necessarily in response
1567 to another message, thes have all been changed to:
1569 Auto-Submitted: auto-replied
1571 in accordance with these statements in the RFC:
1573 The auto-replied keyword:
1575 - SHOULD be used on messages sent in direct response to another
1576 message by an automatic process,
1578 - MUST NOT be used on manually-generated messages,
1580 - MAY be used on Delivery Status Notifications (DSNs) and Message
1581 Disposition Notifications (MDNs),
1583 - MUST NOT be used on messages generated by automatic or periodic
1584 processes, except for messages which are automatic responses to
1587 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1588 to the default Received: header definition.
1590 PH/12 Added log selector acl_warn_skipped (default on).
1592 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1593 variables because (a) they are in the wrong storage pool and (b) even if
1594 they were copied, it wouldn't work properly because of the caching.
1596 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1597 checking when decoding. Apparently there are clients that generate
1598 overlong encoded strings. Why am I not surprised?
1600 PH/15 If the first argument of "${if match_address" was not empty, but did not
1601 contain an "@" character, Exim crashed. Now it writes a panic log message
1602 and treats the condition as false.
1604 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1606 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1607 extension "envelope-auth". The code is finished and in agreement with
1608 other implementations, but there is no documentation so far and in fact,
1609 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1610 not changing the active code.
1612 Print executed "if" and "elsif" statements when debugging is used. This
1613 helps a great deal to understand what a filter does.
1615 Document more things not specified clearly in RFC3028. I had all this
1616 sorted out, when out of a sudden new issues came to my mind. Oops."
1618 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1621 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1622 canonical form (as documented). However, after a host name lookup from
1623 the IP address, check_host() was doing a simple string comparison with
1624 addresses acquired from the DNS when checking that the found name did
1625 have the original IP as one of its addresses. Since any found IPv6
1626 addresses are likely to be in abbreviated form, the comparison could
1627 fail. Luckily, there already exists a function for doing the comparison
1628 by converting both addresses to binary, so now that is used instead of
1629 the text comparison.
1631 PH/20 There was another similar case to PH/19, when a complete host name was
1632 given in a host list; looking up its IP address could give an abbreviated
1633 form, whereas the current host's name might or might not be abbreviated.
1634 The same fix has been applied.
1640 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1641 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1644 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1645 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1647 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1649 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1650 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1651 Also, for "Auto-Submitted", treat anything other than "no" as
1652 non-personal, in accordance with RFC 3834. (Previously it treated
1653 anything starting "auto-" as non-personal.)
1655 TF/01 The control=submission/name=... option had a problem with syntax
1656 errors if the name included a slash character. The /name= option
1657 now slurps the rest of the string, so it can include any characters
1658 but it must come last in the list of options (after /sender_retain
1661 PH/05 Some modifications to the interface to the fake nameserver for the new
1669 TK/01 Added the "success_on_redirect" address verification option. See
1670 NewStuff for rationale and an example.
1672 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1674 PH/02 Patch to exigrep to allow it to work on syslog lines.
1676 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1677 fread() to skip over the body file's header line, because in Cygwin the
1678 header line is locked and is inaccessible.
1680 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1681 co-exist for some time) to make it clear that it is the Exim ID that is
1682 referenced, not the Message-ID: header line.
1684 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1685 string_format() function, because snprintf() does not exist on all
1688 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1689 ${stat: expansion item.
1691 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1692 protocol synchronization error", to keep the pedants happy.
1694 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1695 well as for IRIX systems, when gcc is being used. See the host.c source
1698 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1700 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1703 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1704 transport and to the smtp transport in LMTP mode.
1706 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1708 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1709 run a filter in a subprocess. This could lead to confusion in subsequent
1710 lookups in the parent process. There should also be a search_tidyup() at
1711 the end of the subprocess.
1713 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1714 only if the host matched helo_try_verify_hosts, which caused the
1715 verification to occur when the EHLO/HELO command was issued. The ACL just
1716 tested the remembered result. Now, if a previous verification attempt has
1717 not happened, "verify = helo" does it there and then.
1719 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1721 TK/03 Fix log output including CR from clamd.
1723 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1724 exist provoked a memory error which could cause a segfault.
1726 PH/15 Installed PCRE 6.2
1728 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1730 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1731 of the problem. Specifically, suggested +O2 rather than +O1 for the
1734 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1736 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1737 record, the port was not correctly logged when the outgoing_port log
1738 selector was set (it logged the transort's default port).
1740 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1741 fallback_hosts, and "hosts" in the smtp transport.
1743 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1744 host errors such as "Connection refused".
1746 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1747 authentication with radiusclient 0.4.9:
1749 - Error returned from rc_read_config was caught wrongly
1750 - Username/password not passed on to radius server due to wrong length.
1752 The presumption is that some radiusclient API changes for 4.51/PH/17
1753 were not taken care of correctly. The code is still untested by me (my
1754 Linux distribution still has 0.3.2 of radiusclient), but it was
1755 contributed by a Radius user.
1757 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1758 expanding the "port" option of the smtp transport.
1760 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1761 while reading a MIME header. Thanks to Tom Hughes for a patch.
1763 PH/24 Include config.h inside local_scan.h so that configuration settings are
1766 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1767 that in an ACL for RCPT (for example), you can examine exactly what was
1770 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1771 commands, but it was not correctly comparing the address with the actual
1772 client host address. Thus, it would show the EHLO address in Received:
1773 header lines when this was not necessary.
1775 PH/27 Added the % operator to ${eval:}.
1777 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1778 it should be ignoring failures (because with -C, for example, it has lost
1779 privilege). It wasn't ignoring creation failures other than "already
1782 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1785 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1786 because an input line was too long, either on its own, or by virtue of
1787 too many continuations, the temporary file was not being removed, and the
1788 return code was incorrect.
1790 PH/31 Missing "BOOL" in function definition in filtertest.c.
1792 PH/32 Applied Sieve patches from the maintainer.
1794 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1796 PH/33 Added "verify = not_blind".
1798 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1799 Local/Makefile (with some defaults set). These are used in built scripts
1800 such as exicyclog, but they have never been used in the exim_install
1801 script (though there are many overriding facilities there). I have
1802 arranged that the exim_install script now takes note of these two
1805 PH/35 Installed configuration files for Dragonfly.
1807 PH/36 When a locally submitted message by a trusted user did not contain a
1808 From: header, and the sender address was obtained from -f or from an SMTP
1809 MAIL command, and the trusted user did not use -F to supply a sender
1810 name, $originator_name was incorrectly used when constructing a From:
1811 header. Furthermore, $originator_name was used for submission mode
1812 messages from external hosts without From: headers in a similar way,
1813 which is clearly wrong.
1815 PH/37 Added control=suppress_local_fixups.
1817 PH/38 When log_selector = +received_sender was set, and the addition of the
1818 sender made the log line's construction buffer exactly full, or one byte
1819 less than full, an overflow happened when the terminating "\n" was
1822 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1823 when the result of a list match is failure because a DNS lookup failed.
1825 PH/40 RM_COMMAND is now used in the building process.
1827 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1828 the "build-* directories that it finds.
1830 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1831 address in a domain literal was a prefix of an interface address.
1833 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1834 when verifying a sender address, unless rewrite_headers is false.
1836 PH/44 Wrote a long comment about why errors_to addresses are verified as
1837 recipients, not senders.
1839 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1840 the ratelimit ACL was added.
1842 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1844 PH/46 Added extra information about PostgreSQL errors to the error string.
1846 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1847 avoiding the need to install special zones in a real server. This is
1848 backwards compatible; if it can't find the fake resolver, it drops back.
1849 Thus, both old and new test suites can be run.
1851 TF/02 Added util/ratelimit.pl
1853 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1854 clock is set back in time.
1856 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1857 Candler <B.Candler@pobox.com>.
1859 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1860 OK for addresses that are the result of redirections.
1862 PH/48 A number of further additions for the benefit of the new test suite,
1863 including a fake gethostbyname() that interfaces to the fake DNS resolver
1866 TF/06 The fix for widen_domains has also been applied to qualify_single and
1867 search_parents which are the other dnslookup options that can cause
1870 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1873 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1875 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1876 service" if it encountered an MX record with an empty target hostname.
1877 The message is now "an MX or SRV record indicated no SMTP service".
1879 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1880 if the DNS of the sending site is misconfigured. This is quite a
1881 common situation. This change restores the behaviour of treating a
1882 helo verification defer as a failure.
1884 PH/51 If self=fail was set on a router, the bounce message did not include the
1885 actual error message.
1891 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1893 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1894 command itself ended in error, the underlying message about the transport
1895 filter timeout was being overwritten with the pipe command error. Now the
1896 underlying error message should be appended to the second error message.
1898 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1900 PH/02 Reduce the amount of output that "make" produces by default. Full output
1901 can still be requested.
1903 PH/03 The warning log line about a condition test deferring for a "warn" verb
1904 was being output only once per connection, rather than after each
1905 occurrence (because it was using the same function as for successful
1906 "warn" verbs). This seems wrong, so I have changed it.
1908 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1909 it should not have, which might have caused a crash in the right
1910 circumstances, but probably never did.
1912 PH/04 Installed a modified version of Tony Finch's patch to make submission
1913 mode fix the return path as well as the Sender: header line, and to
1914 add a /name= option so that you can make the user's friendly name appear
1917 TF/03 Added the control = fakedefer ACL modifier.
1919 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1920 Mark Lowes for thorough testing.
1922 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1924 TK/03 Merged latest SRS patch from Miles Wilton.
1926 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1927 with the definition in sysexits.h (which is #included earlier).
1928 Fortunately, Exim does not actually use EX_OK. The code used to try to
1929 preserve the sysexits.h value, by assumimg that macro definitions were
1930 scanned for macro replacements. I have been disabused of this notion,
1931 so now the code just undefines EX_OK before #including unistd.h.
1933 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1934 in the smtp transport. When a block could not be written in a single
1935 write() function, the timeout was being re-applied to each part-write.
1936 This seems wrong - if the receiver was accepting one byte at a time it
1937 would take for ever. The timeout is now adjusted when this happens. It
1938 doesn't have to be particularly precise.
1940 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1941 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1943 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1944 without a domain if the check to <postmaster@domain> fails.
1946 SC/01 Eximstats: added -xls and the ability to specify output files
1947 (patch written by Frank Heydlauf).
1949 SC/02 Eximstats: use FileHandles for outputing results.
1951 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1953 SC/04 Eximstats: fixed display of large numbers with -nvr option
1955 SC/05 Eximstats: fixed merging of reports with empty tables.
1957 SC/06 Eximstats: added the -include_original_destination flag
1959 SC/07 Eximstats: removed tabs and trailing whitespace.
1961 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1963 TK/06 MBOX spool code: Add real "From " MBOX separator line
1964 so the .eml file is really in mbox format (even though
1965 most programs do not really care). Patch from Alex Miller.
1967 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1968 The latter is generated from $received_to and is only set if the
1969 message has one envelope recipient. SA can use these headers,
1970 obviously out-of-the-box. Patch from Alex Miller.
1972 PH/08 The ${def test on a variable was returning false if the variable's
1973 value was "0", contrary to what the specification has always said!
1974 The result should be true unless the variable is empty.
1976 PH/09 The syntax error of a character other than { following "${if
1977 def:variable_name" (after optional whitespace) was not being diagnosed.
1978 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1979 accidental colon was present, for example, could give incorrect results.
1981 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1982 result is used (not including appendfile, where other changes are about
1985 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1986 This involved changing a lot of size variables from int to off_t. It
1987 should work with maildirs and everything.
1989 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1990 spamd dying while we are connected to it.
1992 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1995 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1996 function for BDB 4.3.
1998 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2000 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2001 the sources for PCRE have been re-arranged and more files are now
2004 PH/15 The code I had for printing potentially long long variables in PH/11
2005 above was not the best (it lost precision). The length of off_t variables
2006 is now inspected at build time, and an appropriate printing format (%ld
2007 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2008 to be "long long int" or "long int". This is needed for the internal
2009 formatting function string_vformat().
2011 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2012 the configuration file to be ":syslog", then the script "guesses" where
2013 the logs files are, rather than using the compiled in default. In our
2014 case the guess is not the same as the compiled default, so the script
2015 suddenly stopped working when I started to use syslog. The patch checks
2016 to see if log_file_path is "". If so, it attempts to read it from exim
2017 with no configuration file to get the compiled in version, before it
2018 falls back to the previous guessing code."
2020 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2021 implementing BATV in an Exim configuration. See NewStuff for the gory
2024 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2025 Makefile that are specific to HP-UX.
2027 PH/18 If the "use_postmaster" option was set for a recipient callout together
2028 with the "random" option, the postmaster address was used as the MAIL
2029 FROM address for the random test, but not for the subsequent recipient
2030 test. It is now used for both.
2032 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2033 patch removes a few documentation additions to RFC 3028, because the
2034 latest draft now contains them. It adds the new en;ascii-case comparator
2035 and a new error check for 8bit text in MIME parts. Comparator and
2036 require names are now matched exactly. I enabled the subaddress
2037 extension, but it is not well tested yet (read: it works for me)."
2039 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2040 rework some of the code of TK/09 above to avoid the hardwired use of
2041 "%lld" and "long long". Replaced the call to snprintf() with a call to
2044 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2045 records point to non-existent hosts", "retry timeout exceeded", and
2046 "retry time not reached for any host after a long failure period".
2048 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2049 experimental DomainKeys support:
2051 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2052 (2) On an error such as an illegally used "control", the wrong name for
2053 the control was given.
2055 These problems did NOT occur unless DomainKeys support was compiled.
2057 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2059 PH/24 Added ${if match_ip condition.
2061 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2062 databases so that it will be absolutely obvious if a crash occurs in the
2063 DB library. This is a regular occurrence (often caused by mis-matched
2066 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2067 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2068 picked up on a user's system that detects such things. There doesn't seem
2069 to be a gcc warning option for this - only an attribute that has to be
2070 put on the function's prototype. It seems that in Fedora Core 4 they have
2071 set this on a number of new functions. No doubt there will be more in due
2074 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2075 specify a transport. However, if an address that was verified by such a
2076 router was the subject of a callout, Exim crashed because it tried to
2077 read the rcpt_include_affixes from the non-existent transport. Now it
2078 just assumes that the setting of that option is false. This bug was
2079 introduced by 4.51/PH/31.
2081 PH/28 Changed -d+all to exclude +memory, because that information is very
2082 rarely of interest, but it makes the output a lot bigger. People tend to
2083 do -d+all out of habit.
2085 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2086 code in os-type was giving problems when libc.so lives in lib64, like on
2089 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2090 aren't the modern standard, and it seems that some systems' include files
2091 don't always have them. Exim was already checking for some of the newer
2092 ones like T_AAAA, and defining it itself. I've added checks for all the
2093 record types that Exim uses.
2095 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2096 not automatically generating a new one, as it is supposed to. This
2097 prevented TLS from working. If the file did exist, but contained invalid
2098 data, a new version was generated, as expected. It was only the case of a
2099 non-existent file that was broken.
2101 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2102 with a change in libdomainkeys > 0.64.
2104 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2105 from DNS. If the selector record carries the flag, it now has
2106 precedence over the domain-wide flag.
2108 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2110 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2111 the use of a port name that isn't defined in /etc/services) occurred, the
2112 message was deferred as in a normal delivery, and thus remained on the
2113 spool, instead of being failed because of the mua_wrapper setting. This
2114 is now fixed, and I tidied up some of the mua_wrapper messages at the
2117 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2118 the messages in a hash of arrays rather than using individual hashes.
2119 This is a bit cleaner and results in dramatic memory savings, albeit
2120 at a slight CPU cost.
2122 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2123 as requested by Marc Sherman.
2125 SC/10 Eximstats: added histograms for user specified patterns as requested
2128 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2130 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2131 fopen() in the content-scanning modules that did not already have it.
2137 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2138 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2140 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2142 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2144 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2145 to test Sieve filters that use "vacation".
2147 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2148 that changes the way the GnuTLS parameters are stored in the cache file.
2149 The new format can be generated externally. For backward compatibility,
2150 if the data in the cache doesn't make sense, Exim assumes it has read an
2151 old-format file, and it generates new data and writes a new file. This
2152 means that you can't go back to an older release without removing the
2155 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2156 work if there are any delivery delays because "one_time" forces the
2157 parent to be marked "delivered", so its unseen clone is never tried
2158 again. For this reason, Exim now forbids the simultaneous setting of
2161 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2162 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2163 entry if you want to know the details. The fix, however, neglected to
2164 consider the case when local delivery batching is involved. The test for
2165 "previously delivered" was not happening when checking to see if an
2166 address could be batched with a previous (undelivered) one; under
2167 certain circumstances this could lead to multiple deliveries to the same
2170 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2171 in its include files, and this causes problems building Exim.
2173 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2174 header_syntax) but Exim was just ignoring anything given after a slash.
2175 In particular, this caused confusion with an attempt to use "verify =
2176 reverse_host_lookup/defer_ok". An error is now given when options are
2177 supplied for verify items that do not have them. (Maybe reverse_host_
2178 lookup should have a defer_ok option, but that's a different point.)
2180 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2181 defined by RFC 821) to 2048, because there were problems with some AUTH
2182 commands, and RFC 1869 says the size should be increased for extended
2183 SMTP commands that take arguments.
2185 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2188 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2189 "unknown" error; now it says that the functionality isn't in the binary.
2191 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2192 an address' error message when a string expansion fails (syntax or
2193 whatever). Otherwise the password may appear in the log. Following change
2194 PH/42 below, there is no longer a chance of it appearing in a bounce
2197 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2199 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2200 of its addresses was ever tried. (Bugzilla bug #2.)
2202 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2203 the result incorrectly in the debug output. (It correctly added a newline
2204 to what was transported.)
2206 TF/01 Added $received_time.
2208 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2209 commented out examples of how to interface to a virus scanner and to
2210 SpamAssassin. Also added commented examples of av_scanner and
2211 spamd_address settings.
2213 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2214 and controls are allowed in which ACLs. There were a couple of minor
2215 errors. Some of the entries in the conditions table (which is a table of
2216 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2217 negation of where the condition IS allowed.
2219 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2221 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2222 header file does not have a version number, so I've had to invent a new
2223 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2224 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2225 radiusclient), but it was contributed by a Radius user.
2227 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2228 files or directories, for OpenSSL.
2230 PH/19 When an Exim process that is running as root has to create an Exim log
2231 file, it does so in a subprocess that runs as exim:exim so as to get the
2232 ownership right at creation (otherwise, other Exim processes might see
2233 the file with the wrong ownership). There was no test for failure of this
2234 fork() call, which would lead to the process getting stuck as it waited
2235 for a non-existent subprocess. Forks do occasionally fail when resources
2236 run out. I reviewed all the other calls to fork(); they all seem to check
2239 PH/20 When checking for unexpected SMTP input at connect time (before writing
2240 the banner), Exim was not dealing correctly with a non-positive return
2241 from the read() function. If the client had disconnected by this time,
2242 the result was a log entry for a synchronization error with an empty
2243 string after "input=" when read() returned zero. If read() returned -1
2244 (an event I could not check), uninitialized data bytes were printed.
2245 There were reports of junk text (parts of files, etc) appearing after
2248 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2250 PH/22 Added support for macro redefinition, and (re)definition in between
2251 driver and ACL definitions.
2253 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2254 forgetting to use the resulting value; it was using the unexpanded value.
2256 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2257 hadn't been configured. The fix is from Juergen Kreileder, who
2258 understands it better than I do:
2260 "Here's what I see happening with three configured cyrus_sasl
2261 authenticators configured (plain, login, cram-md5):
2263 On startup auth_cyrus_sasl_init() gets called for each of these.
2264 This means three calls to sasl_listmech() without a specified mech_list.
2265 => SASL tests which mechs of all available mechs actually work
2266 => three warnings about OTP not working
2267 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2269 With the patch, sasl_listmech() also gets called three times. But now
2270 SASL's mech_list option is set to the server_mech specified in the the
2271 authenticator. Or in other words, the answer from sasl_listmech()
2272 gets limited to just the mech you're testing for (which is different
2274 => the return list contains just 'plain' or 'login', 'cram-md5' or
2275 nothing depending on the value of ob->server_mech.
2277 I've just tested the patch: Authentication still works fine,
2278 unavailable mechs specified in the exim configuration are still
2279 caught, and the auth.log warnings about OTP are gone."
2281 PH/25 When debugging is enabled, the contents of the command line are added
2282 to the debugging output, even when log_selector=+arguments is not
2285 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2286 answer is "GNU", and only if the return is "GNU/something" is the answer
2289 PH/27 $acl_verify_message is now set immediately after the failure of a
2290 verification in an ACL, and so is available in subsequent modifiers. In
2291 particular, the message can be preserved by coding like this:
2293 warn !verify = sender
2294 set acl_m0 = $acl_verify_message
2296 Previously, $acl_verify_message was set only while expanding "message"
2297 and "log_message" when a very denied access.
2299 PH/28 Modified OS/os.c-Linux with
2301 -#ifndef OS_LOAD_AVERAGE
2302 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2304 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2305 nomenclature these days.)
2307 PH/29 Installed patch from the Sieve maintainer that adds the options
2308 sieve_useraddress and sieve_subaddress to the redirect router.
2310 PH/30 In these circumstances:
2311 . Two addresses routed to the same list of hosts;
2312 . First host does not offer TLS;
2313 . First host accepts first address;
2314 . First host gives temporary error to second address;
2315 . Second host offers TLS and a TLS session is established;
2316 . Second host accepts second address.
2317 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2318 and peerdn, if requested) that were in fact used only for the second
2321 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2322 attention to any local part prefix or suffix that was matched by the
2323 router that accepted the address. It now behaves in the same way as it
2324 does for delivery: the affixes are removed from the local part unless
2325 rcpt_include_affixes is set on the transport.
2327 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2328 timeout during the DATA phase of an incoming message.
2330 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2331 applied a patch sanctioned by the Sieve maintainer.
2333 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2334 the uid or gid is negative. A case of a negative gid caused this to be
2335 noticed. The fix allows for either to be negative.
2337 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2338 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2341 PH/36 The change PH/12 above was broken. Fixed it.
2343 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2344 the grounds that routing the same address twice would always produce the
2345 same answer. This might have been true once, but it is certainly no
2346 longer true now. Routing a child address may depend on the previous
2347 routing that produced that child. Some complicated redirection strategies
2348 went wrong when messages had multiple recipients, and made Exim's
2349 behaviour dependent on the order in which the addresses were given.
2351 I have moved the duplicate checking until after the routing is complete.
2352 Exim scans the addresses that are assigned to local and remote
2353 transports, and removes any duplicates. This means that more work will be
2354 done, as duplicates will always all be routed, but duplicates are
2355 presumably rare, so I don't expect this is of any significance.
2357 For deliveries to pipes, files, and autoreplies, the duplicate checking
2358 still happens during the routing process, since they are not going to be
2361 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2362 It corrects a timeout issue with spamd. This is Ian's comment: "The
2363 background is that sometimes spamd either never reads data from a
2364 connection it has accepted, or it never writes response data. The exiscan
2365 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2366 blindly assumes that writes won't block so it may never time out."
2368 PH/39 Allow G after quota size as well as K and M.
2370 PH/40 The value set for $authenticated_id in an authenticator may not contain
2371 binary zeroes or newlines because the value is written to log lines and
2372 to spool files. There was no check on this. Now the value is run through
2373 the string_printing() function so that such characters are converted to
2374 printable escape sequences.
2376 PH/41 $message_linecount is a new variable that contains the total number of
2377 lines in the message. Compare $body_linecount, which is the count for the
2380 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2381 bounce and delay warning messages, except in certain special cases, which
2384 (a) An SMTP error message from a remote host;
2385 (b) A message specified in a :fail: redirection;
2386 (c) A message specified in a "fail" command in a system filter;
2387 (d) A message specified in a FAIL return from the queryprogram router;
2388 (e) A message specified by the cannot_route_message router option.
2390 In these cases only, Exim does include the error details in bounce and
2391 warning messages. There are also a few cases where bland messages such
2392 as "unrouteable address" or "local delivery error" are given.
2394 PH/43 $value is now also set for the "else" part of a ${run expansion.
2396 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2397 being worked on, but at least Exim now implements the latest version to
2400 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2401 process to complete was treated as a delivery failure, a timeout while
2402 writing the message to the pipe was logged, but erroneously treated as a
2403 successful delivery. Such timeouts include transport filter timeouts. For
2404 consistency with the overall process timeout, these timeouts are now
2405 treated as errors, giving rise to delivery failures by default. However,
2406 there is now a new Boolean option for the pipe transport called
2407 timeout_defer, which, if set TRUE, converts the failures into defers for
2408 both kinds of timeout. A transport filter timeout is now identified in
2411 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2412 systems where "make" and "gmake" are different, calling "gmake" at top
2413 level broke things. I've arranged for the value of $(MAKE) to be passed
2414 from the Makefile to this script so that it can call the same version of
2418 A note about Exim versions 4.44 and 4.50
2419 ----------------------------------------
2421 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2422 changes of various kinds. As a consequence, a big documentation update was
2423 needed. This delayed the release for rather longer than seemed good, especially
2424 in the light of a couple of (minor) security issues. Therefore, the changes
2425 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2426 release. So 4.44 and 4.50 are in effect two different branches that both start
2429 I have left the 4.50 change log unchanged; it contains all the changes since
2430 4.43. The change log for 4.44 is below; many of its items are identical to
2431 those for 4.50. This seems to be the most sensible way to preserve the
2432 historical information.
2438 1. Minor wording change to the doc/README.SIEVE file.
2440 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2441 computation of the current number of files was incorrect.
2443 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2444 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2447 4. Give more explanation in the error message when the command for a transport
2448 filter fails to execute.
2450 5. There are several places where Exim runs a non-Exim command in a
2451 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2452 was being done only for the command run by the queryprogram router. It is
2453 now done for all such subprocesses. The other cases are: ${run, transport
2454 filters, and the commands run by the lmtp and pipe transports.
2456 6. Added CONFIGURE_GROUP build-time option.
2458 7. Some older OS have a limit of 256 on the maximum number of file
2459 descriptors. Exim was using setrlimit() to set 1000 as a large value
2460 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2461 systems. I've change it so that if it can't get 1000, it tries for 256.
2463 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2464 was an oversight, and furthermore, ever since the addition of extra
2465 controls (e.g. 4.43/32), the checks on when to allow different forms of
2466 "control" were broken. There should now be diagnostics for all cases when a
2467 control that does not make sense is encountered.
2469 9. Added the /retain_sender option to "control=submission".
2471 10. $recipients is now available in the predata ACL (oversight).
2473 11. Tidy the search cache before the fork to do a delivery from a message
2474 received from the command line. Otherwise the child will trigger a lookup
2475 failure and thereby defer the delivery if it tries to use (for example) a
2476 cached ldap connection that the parent has called unbind on.
2478 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2479 of $address_data from the recipient verification was clobbered by the
2480 sender verification.
2482 13. The value of address_data from a sender verification is now available in
2483 $sender_address_data in subsequent conditions in the ACL statement.
2485 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2487 15. Added a new option "connect=<time>" to callout options, to set a different
2490 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2491 was its contents. (It was OK if the option was not defined at all.)
2493 17. A "Completed" log line is now written for messages that are removed from
2494 the spool by the -Mrm option.
2496 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2497 information about exactly what failed.
2499 19. Added -dd to debug only the daemon process.
2501 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2502 handles timeouts, both on the server side and network timeouts. Renamed the
2503 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2505 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2506 It is now set to "smtps".
2508 22. $host_address is now set to the target address during the checking of
2509 ignore_target_hosts.
2511 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2512 being passed; this would have caused $sender_host_name to have been used if
2513 matching the list had actually called for a host name (not very likely,
2514 since this list is usually IP addresses). A host name is now passed as
2517 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2518 code that specifies a non-restarting handler (typically sigaction() in
2519 modern systems) in an attempt to fix a rare and obscure crash bug.
2521 25. Narrowed the window for a race in the daemon that could cause it to ignore
2522 SIGCHLD signals. This is not a major problem, because they are used only to
2523 wake it up if nothing else does.
2525 26. A malformed maildirsize file could cause Exim to calculate negative values
2526 for the mailbox size or file count. Odd effects could occur as a result.
2527 The maildirsize information is now recalculated if the size or filecount
2530 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2531 support for a long time. Removed HAVE_SYS_VFS_H.
2533 28. Installed the latest version of exipick from John Jetmore.
2535 29. In an address list, if the pattern was not a regular expression, an empty
2536 subject address (from a bounce message) matched only if the pattern was an
2537 empty string. Non-empty patterns were not even tested. This was the wrong
2538 because it is perfectly reasonable to use an empty address as part of a
2539 database query. An empty address is now tested by patterns that are
2540 lookups. However, all the other forms of pattern expect the subject to
2541 contain a local part and a domain, and therefore, for them, an empty
2542 address still always fails if the pattern is not itself empty.
2544 30. Exim went into a mad DNS loop when attempting to do a callout where the
2545 host was specified on an smtp transport, and looking it up yielded more
2546 than one IP address.
2548 31. Re-factored the code for checking spool and log partition space into a
2549 function that finds that data and another that does the check. The former
2550 is then used to implement four new variables: $spool_space, $log_space,
2551 $spool_inodes, and $log_inodes.
2553 32. The RFC2047 encoding function was originally intended for short strings
2554 such as real names; it was not keeping to the 75-character limit for
2555 encoded words that the RFC imposes. It now respects the limit, and
2556 generates multiple encoded words if necessary. To be on the safe side, I
2557 have increased the buffer size for the ${rfc2047: expansion operator from
2560 33. It is now permitted to omit both strings after an "if" condition; if the
2561 condition is true, the result is "true". As before, when the second string
2562 is omitted, a false condition yields an empty string. This makes it less
2563 cumbersome to write custom ACL and router conditions.
2565 34. Failure to deliver a bounce message always caused it to be frozen, even if
2566 there was an errors_to setting on the router. The errors_to setting is now
2569 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2570 canonical form (fully expanded) before being placed in
2571 $sender_host_address.
2573 36. The table in the code that translates DNS record types into text (T_A to
2574 "A" for instance) was missing entries for NS and CNAME. It is just possible
2575 that this could have caused confusion if both these types were looked up
2576 for the same domain, because the text type is used as part of Exim's
2577 per-process caching. But the chance of anyone hitting this buglet seems
2580 37. The dnsdb lookup has been extended in a number of ways.
2582 (1) There is a new type, "zns", which walks up the domain tree until it
2583 finds some nameserver records. It should be used with care.
2585 (2) There is a new type, "mxh", which is like "mx" except that it returns
2586 just the host names, not the priorities.
2588 (3) It is now possible to give a list of domains (or IP addresses) to be
2589 looked up. The behaviour when one of the lookups defers can be
2590 controlled by a keyword.
2592 (4) It is now possible to specify the separator character for use when
2593 multiple records are returned.
2595 38. The dnslists ACL condition has been extended: it is now possible to supply
2596 a list of IP addresses and/or domains to be looked up in a particular DNS
2599 39. Added log_selector=+queue_time_overall.
2601 40. When running the queue in the test harness, wait just a tad after forking a
2602 delivery process, to get repeatability of debugging output.
2604 41. Include certificate and key file names in error message when GnuTLS fails
2605 to set them up, because the GnuTLS error message doesn't include the name
2606 of the failing file when there is a problem reading it.
2608 42. Allow both -bf and -bF in the same test run.
2610 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2612 44. The "Exiscan patch" is now merged into the mainline Exim source.
2614 45. Sometimes the final signoff response after QUIT could fail to get
2615 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2616 < 0 before doing a fflush(). This bug looks as though it goes back to the
2617 introduction of TLS in release 3.20, but "sometimes" must have been rare
2618 because the tests only now provoked it.
2620 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2621 (this can affect the format of dates).
2623 47. exim_tidydb, when checking for the continued existence of a message for
2624 which it has found a message-specific retry record, was not finding
2625 messages that were in split spool directories. Consequently, it was
2626 deleting retry records that should have stayed in existence.
2628 48. Steve fixed some bugs in eximstats.
2630 49. The SPA authentication driver was not abandoning authentication and moving
2631 on to the next authenticator when an expansion was forced to fail,
2632 contradicting the general specification for all authenticators. Instead it
2633 was generating a temporary error. It now behaves as specified.
2635 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2636 (the order specifies the preference for clients). The order is now AES256,
2637 AES128, 3DES, ARCFOUR128.
2639 51. Small patch to Sieve code - explicitly set From: when generating an
2642 52. Exim crashed if a remote delivery caused a very long error message to be
2643 recorded - for instance if somebody sent an entire SpamAssassin report back
2644 as a large number of 550 error lines. This bug was coincidentally fixed by
2645 increasing the size of one of Exim's internal buffers (big_buffer) that
2646 happened as part of the Exiscan merge. However, to be on the safe side, I
2647 have made the code more robust (and fixed the comments that describe what
2650 53. Now that there can be additional text after "Completed" in log lines (if
2651 the queue_time_overall log selector is set), a one-byte patch to exigrep
2652 was needed to allow it to recognize "Completed" as not the last thing in
2655 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2656 patch that reportedly fixes this has been added. I am not expert enough to
2657 create a test for it. This is what the patch creator wrote:
2659 "I found a little strange behaviour of ldap code when working with
2660 Windows 2003 AD Domain, where users was placed in more than one
2661 Organization Units. When I tried to give exim partial DN, the exit code
2662 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2663 But simultaneously result of request was absolutely normal ldap result,
2664 so I produce this patch..."
2666 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2667 REFERENCE, so I have modified the code to exclude the patch when that macro
2670 55. Some experimental protocols are using DNS PTR records for new purposes. The
2671 keys for these records are domain names, not reversed IP addresses. The
2672 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2673 leaves it alone. Component reversal etc. now happens only for IP addresses.
2676 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2678 57. Double the size of the debug message buffer (to 2048) so that more of very
2679 long debug lines gets shown.
2681 58. The exicyclog utility now does better if the number of log files to keep
2682 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2684 59. Two changes related to the smtp_active_hostname option:
2686 (1) $smtp_active_hostname is now available as a variable.
2687 (2) The default for smtp_banner uses $smtp_active_hostname instead
2688 of $primary_hostname.
2690 60. The host_aton() function is supposed to be passed a string that is known
2691 to be a valid IP address. However, in the case of IPv6 addresses, it was
2692 not checking this. This is a hostage to fortune. Exim now panics and dies
2693 if the condition is not met. A case was found where this could be provoked
2694 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2695 components; fortuitously, this particular loophole had already been fixed
2696 by change 4.50/55 above.
2698 If there are any other similar loopholes, the new check in host_aton()
2699 itself should stop them being exploited. The report I received stated that
2700 data on the command line could provoke the exploit when Exim was running as
2701 exim, but did not say which command line option was involved. All I could
2702 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2703 running as the user.
2706 61. There was a buffer overflow vulnerability in the SPA authentication code
2707 (which came originally from the Samba project). I have added a test to the
2708 spa_base64_to_bits() function which I hope fixes it.
2711 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2712 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2714 63. The daemon start-up calls getloadavg() while still root for those OS that
2715 need the first call to be done as root, but it missed one case: when
2716 deliver_queue_load_max is set with deliver_drop_privilege. This is
2717 necessary for the benefit of the queue runner, because there is no re-exec
2718 when deliver_drop_privilege is set.
2720 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2721 This has been fixed.
2723 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2724 was in use, was not putting the data itself into the right store pool;
2725 consequently, it could be overwritten for a subsequent message in the same
2726 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2729 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2731 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2732 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2733 the function were treating the return as a boolean value, which happened to
2734 work because 0=false and not-0=true, but is not correct code.
2736 68. The host_aton() function was not handling scoped IPv6 addresses (those
2737 with, for example, "%eth0" on the end) correctly.
2739 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2740 negated items (that is, ~something) in unsigned ints. Some compilers
2741 apparently mutter when there is no cast.
2743 70. If an address verification called from an ACL failed, and did not produce a
2744 user-specific message (i.e. there was only a "system" message), nothing was
2745 put in $acl_verify_message. In this situation, it now puts the system
2748 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2749 session; change 4.31/43 added the unwanted input to the log line - except
2750 that it did not do this in the start of session case. It now does.
2752 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2753 This is wrong and can cause the other end to generate a synchronization
2754 error if it is another Exim or anything else that does the synchronization
2755 check. A QUIT command is no longer sent after a timeout.
2757 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2758 during host lookups.
2760 74. The defer_ok option of callout verification was not working if it was used
2761 when verifying addresses in header lines, that is, for this case:
2763 verify = header_sender/callout=defer_ok
2765 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2766 those file descriptors could be used for SMTP connections. If anything
2767 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2768 could be sent to the SMTP client, causing chaos. The daemon now opens
2769 stdin, stdout, and stderr to /dev/null when it puts itself into the
2772 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2773 log by default. The user can override this with suitable Perl magic.
2775 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2776 the log message when discard triggers, was not working for the DATA ACL or
2777 for the non-SMTP ACL.
2779 78. Error message wording change in sieve.c.
2781 79. If smtp_accept_max_per_host was set, the number of connections could be
2782 restricted to fewer than expected, because the daemon was trying to set up
2783 a new connection before checking whether the processes handling previous
2784 connections had finished. The check for completed processes is now done
2785 earlier. On busy systems, this bug wouldn't be noticed because something
2786 else would have woken the daemon, and it would have reaped the completed
2789 80. If a message was submitted locally by a user whose login name contained one
2790 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2791 It caused a spool format error. I have fixed the spool reading code. A
2792 related problem was that the "from" clause in the Received: line became
2793 illegal because of the space(s). It is now covered by ${quote_local_part.
2795 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2798 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2800 83. Patch from Sieve maintainer to fix unterminated string problem in
2801 "vacation" handling.
2803 84. Some minor changes to the Linux configuration files to help with other
2804 OS variants using glibc.
2806 85. One more patch for Sieve to update vacation handling to latest spec.
2809 ----------------------------------------------------
2810 See the note above about the 4.44 and 4.50 releases.
2811 ----------------------------------------------------
2817 1. Change 4.43/35 introduced a bug that caused file counts to be
2818 incorrectly computed when quota_filecount was set in an appendfile
2821 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2822 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2825 3. Give more explanation in the error message when the command for a transport
2826 filter fails to execute.
2828 4. There are several places where Exim runs a non-Exim command in a
2829 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2830 was being done only for the command run by the queryprogram router. It is
2831 now done for all such subprocesses. The other cases are: ${run, transport
2832 filters, and the commands run by the lmtp and pipe transports.
2834 5. Some older OS have a limit of 256 on the maximum number of file
2835 descriptors. Exim was using setrlimit() to set 1000 as a large value
2836 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2837 systems. I've change it so that if it can't get 1000, it tries for 256.
2839 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2840 was an oversight, and furthermore, ever since the addition of extra
2841 controls (e.g. 4.43/32), the checks on when to allow different forms of
2842 "control" were broken. There should now be diagnostics for all cases when a
2843 control that does not make sense is encountered.
2845 7. $recipients is now available in the predata ACL (oversight).
2847 8. Tidy the search cache before the fork to do a delivery from a message
2848 received from the command line. Otherwise the child will trigger a lookup
2849 failure and thereby defer the delivery if it tries to use (for example) a
2850 cached ldap connection that the parent has called unbind on.
2852 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2853 of $address_data from the recipient verification was clobbered by the
2854 sender verification.
2856 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2857 was its contents. (It was OK if the option was not defined at all.)
2859 11. A "Completed" log line is now written for messages that are removed from
2860 the spool by the -Mrm option.
2862 12. $host_address is now set to the target address during the checking of
2863 ignore_target_hosts.
2865 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2866 being passed; this would have caused $sender_host_name to have been used if
2867 matching the list had actually called for a host name (not very likely,
2868 since this list is usually IP addresses). A host name is now passed as
2871 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2872 code that specifies a non-restarting handler (typically sigaction() in
2873 modern systems) in an attempt to fix a rare and obscure crash bug.
2875 15. Narrowed the window for a race in the daemon that could cause it to ignore
2876 SIGCHLD signals. This is not a major problem, because they are used only to
2877 wake it up if nothing else does.
2879 16. A malformed maildirsize file could cause Exim to calculate negative values
2880 for the mailbox size or file count. Odd effects could occur as a result.
2881 The maildirsize information is now recalculated if the size or filecount
2884 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2885 support for a long time. Removed HAVE_SYS_VFS_H.
2887 18. Updated exipick to current release from John Jetmore.
2889 19. Allow an empty sender to be matched against a lookup in an address list.
2890 Previously the only cases considered were a regular expression, or an
2893 20. Exim went into a mad DNS lookup loop when doing a callout where the
2894 host was specified on the transport, if the DNS lookup yielded more than
2897 21. The RFC2047 encoding function was originally intended for short strings
2898 such as real names; it was not keeping to the 75-character limit for
2899 encoded words that the RFC imposes. It now respects the limit, and
2900 generates multiple encoded words if necessary. To be on the safe side, I
2901 have increased the buffer size for the ${rfc2047: expansion operator from
2904 22. Failure to deliver a bounce message always caused it to be frozen, even if
2905 there was an errors_to setting on the router. The errors_to setting is now
2908 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2909 canonical form (fully expanded) before being placed in
2910 $sender_host_address.
2912 24. Updated eximstats to version 1.33
2914 25. Include certificate and key file names in error message when GnuTLS fails
2915 to set them up, because the GnuTLS error message doesn't include the name
2916 of the failing file when there is a problem reading it.
2918 26. Expand error message when OpenSSL has problems setting up cert/key files.
2921 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2922 (this can affect the format of dates).
2924 28. exim_tidydb, when checking for the continued existence of a message for
2925 which it has found a message-specific retry record, was not finding
2926 messages that were in split spool directories. Consequently, it was
2927 deleting retry records that should have stayed in existence.
2929 29. eximstats updated to version 1.35
2930 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2931 1.35 - bugfix such that pie charts by volume are generated correctly
2933 30. The SPA authentication driver was not abandoning authentication and moving
2934 on to the next authenticator when an expansion was forced to fail,
2935 contradicting the general specification for all authenticators. Instead it
2936 was generating a temporary error. It now behaves as specified.
2938 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2939 (the order specifies the preference for clients). The order is now AES256,
2940 AES128, 3DES, ARCFOUR128.
2942 31. Small patch to Sieve code - explicitly set From: when generating an
2945 32. Exim crashed if a remote delivery caused a very long error message to be
2946 recorded - for instance if somebody sent an entire SpamAssassin report back
2947 as a large number of 550 error lines. This bug was coincidentally fixed by
2948 increasing the size of one of Exim's internal buffers (big_buffer) that
2949 happened as part of the Exiscan merge. However, to be on the safe side, I
2950 have made the code more robust (and fixed the comments that describe what
2953 33. Some experimental protocols are using DNS PTR records for new purposes. The
2954 keys for these records are domain names, not reversed IP addresses. The
2955 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2956 leaves it alone. Component reversal etc. now happens only for IP addresses.
2959 34. The host_aton() function is supposed to be passed a string that is known
2960 to be a valid IP address. However, in the case of IPv6 addresses, it was
2961 not checking this. This is a hostage to fortune. Exim now panics and dies
2962 if the condition is not met. A case was found where this could be provoked
2963 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2964 components; fortuitously, this particular loophole had already been fixed
2965 by change 4.50/55 or 4.44/33 above.
2967 If there are any other similar loopholes, the new check in host_aton()
2968 itself should stop them being exploited. The report I received stated that
2969 data on the command line could provoke the exploit when Exim was running as
2970 exim, but did not say which command line option was involved. All I could
2971 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2972 running as the user.
2975 35. There was a buffer overflow vulnerability in the SPA authentication code
2976 (which came originally from the Samba project). I have added a test to the
2977 spa_base64_to_bits() function which I hope fixes it.
2980 36. The daemon start-up calls getloadavg() while still root for those OS that
2981 need the first call to be done as root, but it missed one case: when
2982 deliver_queue_load_max is set with deliver_drop_privilege. This is
2983 necessary for the benefit of the queue runner, because there is no re-exec
2984 when deliver_drop_privilege is set.
2986 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2987 was in use, was not putting the data itself into the right store pool;
2988 consequently, it could be overwritten for a subsequent message in the same
2989 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2992 38. Sometimes the final signoff response after QUIT could fail to get
2993 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2994 < 0 before doing a fflush(). This bug looks as though it goes back to the
2995 introduction of TLS in release 3.20, but "sometimes" must have been rare
2996 because the tests only now provoked it.
3002 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3003 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3004 Now it returns an integer. A number of calls were still expecting a T/F
3005 return. Fortuitously, in all cases, the tests worked in OK situations,
3006 which is the norm. However, things would have gone wrong on any write
3007 failures on the smtp file descriptor. This function is used when sending
3008 messages over SMTP and also when doing verify callouts.
3010 2. When Exim is called to do synchronous delivery of a locally submitted
3011 message (the -odf or -odi options), it no longer closes stderr before doing
3014 3. Implemented the mua_wrapper option.
3016 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3018 5. Implemented the functions header_remove(), header_testname(),
3019 header_add_at_position(), and receive_remove_recipient(), and exported them
3022 6. If an ACL "warn" statement specified the addition of headers, Exim already
3023 inserted X-ACL-Warn: at the start if there was no header name. However, it
3024 was not making this test for the second and subsequent header lines if
3025 there were newlines in the string. This meant that an invalid header could
3026 be inserted if Exim was badly configured.
3028 7. Allow an ACL "warn" statement to add header lines at the start or after all
3029 the Received: headers, as well as at the end.
3031 8. Added the rcpt_4xx retry error code.
3033 9. Added postmaster_mailfrom=xxx to callout verification option.
3035 10. Added mailfrom=xxxx to the callout verification option, for verify=
3038 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3039 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3041 12. Inserted some casts to stop certain compilers warning when using pointer
3042 differences as field lengths or precisions in printf-type calls (mostly
3043 affecting debugging statements).
3045 13. Added optional readline() support for -be (dynamically loaded).
3047 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3048 same clock tick as a message's arrival, so that its received time was the
3049 same as the "first fail" time on the retry record, and that message
3050 remained on the queue past the ultimate address timeout, every queue runner
3051 would try a delivery (because it was past the ultimate address timeout) but
3052 after another failure, the ultimate address timeout, which should have then
3053 bounced the address, did not kick in. This was a "< instead of <=" error;
3054 in most cases the first failure would have been in the next clock tick
3055 after the received time, and all would be well.
3057 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3058 being recognized when the domain list was tested by the match_domain
3059 condition in an expansion string.
3061 16. Added the ${str2b64: operator.
3063 17. Exim was always calling setrlimit() to set a large limit for the number of
3064 processes, without checking whether the existing limit was already
3065 adequate. (It did check for the limit on file descriptors.) Furthermore,
3066 errors from getrlimit() and setrlimit() were being ignored. Now they are
3067 logged to the main and panic logs, but Exim does carry on, to try to do its
3068 job under whatever limits there are.
3070 18. Imported PCRE 5.0.
3072 19. Trivial typo in log message " temporarily refused connection" (the leading
3075 20. If the log selector return_path_on_delivery was set and an address was
3076 redirected to /dev/null, the delivery process crashed because it assumed
3077 that a return path would always be set for a "successful" delivery. In this
3078 case, the whole delivery is bypassed as an optimization, and therefore no
3081 21. Internal re-arrangement: the function for sending a challenge and reading
3082 a response while authentication was assuming a zero-terminated challenge
3083 string. It's now changed to take a pointer and a length, to allow for
3084 binary data in such strings.
3086 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3088 23. Exim was not respecting finduser_retries when seeking the login of the
3089 uid under which it was called; it was always trying 10 times. (The default
3090 setting of finduser_retries is zero.) Also, it was sleeping after the final
3091 failure, which is pointless.
3093 24. Implemented tls_on_connect_ports.
3095 25. Implemented acl_smtp_predata.
3097 26. If the domain in control=submission is set empty, Exim assumes that the
3098 authenticated id is a complete email address when it generates From: or
3099 Sender: header lines.
3101 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3102 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3103 chown and chgrp in /bin and hostname in /usr/bin.
3105 28. Exim was keeping the "process log" file open after each use, just as it
3106 does for the main log. This opens the possibility of it remaining open for
3107 long periods when the USR1 signal hits a daemon. Occasional processlog
3108 errors were reported, that could have been caused by this. Anyway, it seems
3109 much more sensible not to leave this file open at all, so that is what now
3112 29. The long-running daemon process does not normally write to the log once it
3113 has entered its main loop, and it closes the log before doing so. This is
3114 so that log files can straightforwardly be renamed and moved. However,
3115 there are a couple of unusual error situations where the daemon does write
3116 log entries, and I had neglected to close the log afterwards.
3118 30. The text of an SMTP error response that was received during a remote
3119 delivery was being truncated at 512 bytes. This is too short for some of
3120 the long messages that one sometimes sees. I've increased the limit to
3123 31. It is now possible to make retry rules that apply only when a message has a
3124 specific sender, in particular, an empty sender.
3126 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3127 it possible to be selective about when SMTP synchronization is enforced.
3129 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3131 32. Implemented hosts_connection_nolog.
3133 33. Added an ACL for QUIT.
3135 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3138 35. Added mailbox_size and mailbox_filecount to appendfile.
3140 36. Added control = no_multiline_responses to ACLs.
3142 37. There was a bug in the logic of the code that waits for the clock to tick
3143 in the case where the clock went backwards by a substantial amount such
3144 that the microsecond fraction of "now" was more than the microsecond
3145 fraction of "then" (but the whole seconds number was less).
3147 38. Added support for the libradius Radius client library this is found on
3148 FreeBSD (previously only the radiusclient library was supported).
3154 1. When certain lookups returned multiple values in the form name=value, the
3155 quoting of the values was not always being done properly. Specifically:
3156 (a) If the value started with a double quote, but contained no whitespace,
3158 (b) If the value contained whitespace other than a space character (i.e.
3159 tabs or newlines or carriage returns) it was not quoted.
3160 This fix has been applied to the mysql and pgsql lookups by writing a
3161 separate quoting function and calling it from the lookup code. The fix
3162 should probably also be applied to nisplus, ibase and oracle lookups, but
3163 since I cannot test any of those, I have not disturbed their existing code.
3165 2. A hit in the callout cache for a specific address caused a log line with no
3166 reason for rejecting RCPT. Now it says "Previous (cached) callout
3167 verification failure".
3169 3. There was an off-by-one bug in the queryprogram router. An over-long
3170 return line was truncated at 256 instead of 255 characters, thereby
3171 overflowing its buffer with the terminating zero. As well as fixing this, I
3172 have increased the buffer size to 1024 (and made a note to document this).
3174 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3175 when Exim is waiting for an SMTP response from a remote server, Exim
3176 restarts its select() call on the socket, thereby resetting its timeout.
3177 This is not a problem when such interrupts are rare. Somebody set up a cron
3178 job to run exiwhat every 2 minutes, which is less than the normal select()
3179 timeout (5 or 10 minutes). This meant that the select() timeout never
3180 kicked in because it was always reset. I have fixed this by comparing the
3181 time when an interrupt arrives with the time at the start of the first call
3182 to select(). If more time than the timeout has elapsed, the interrupt is
3183 treated as a timeout.
3185 5. Some internal re-factoring in preparation for the addition of Sieve
3186 extensions (by MH). In particular, the "personal" test is moved to a
3187 separate function, and given an option for scanning Cc: and Bcc: (which is
3188 not set for Exim filters).
3190 6. When Exim created an email address using the login of the caller as the
3191 local part (e.g. when creating a From: or Sender: header line), it was not
3192 quoting the local part when it contained special characters such as @.
3194 7. Installed new OpenBSD configuration files.
3196 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3197 try to make them clearer.
3199 9. Callout options, other than the timeout value, were being ignored when
3200 verifying sender addresses in header lines. For example, when using
3202 verify = header_sender/callout=no_cache
3204 the cache was (incorrectly) being used.
3206 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3207 only the Cygwin environment.
3209 11. When return_path_on_delivery was set as a log selector, if different remote
3210 addresses in the same message used different return paths and parallel
3211 remote delivery occurred, the wrong values would sometimes be logged.
3212 (Whenever a remote delivery process finished, the return path value from
3213 the most recently started remote delivery process was logged.)
3215 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3216 lines when AUTH and/or TLS are in use. This is the "received protocol"
3217 field. Exim used to use "asmtp" for authenticated SMTP, without any
3218 indication (in the protocol name) for TLS use. Now it follows the RFC and
3219 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3220 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3221 names appear in log lines as well as in Received: header lines.
3223 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3224 extensions, and comparison tests, and to fix some bugs.
3226 14. Changes to the "personal" filter test:
3228 (1) The test was buggy in that it was just doing the equivalent of
3229 "contains" tests on header lines. For example, if a user's address was
3230 anne@some.where, the "personal" test would incorrectly be true for
3232 To: susanne@some.where
3234 This test is now done by extracting each address from the header in turn,
3235 and checking the entire address. Other tests that are part of "personal"
3236 are now done using regular expressions (for example, to check local parts
3237 of addresses in From: header lines).
3239 (2) The list of non-personal local parts in From: addresses has been
3240 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3241 taken from the Sieve specification recommendations.
3243 (3) If the message contains any header line starting with "List-" it is
3244 treated as non-personal.
3246 (4) The test for "circular" in the Subject: header line has been removed
3247 because it now seems ill-conceived.
3249 15. Minor typos in src/EDITME comments corrected.
3251 16. Installed latest exipick from John Jetmore.
3253 17. If headers_add on a router specified a text string that was too long for
3254 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3255 of string_sprintf() is now avoided.
3257 18. $message_body_size was not set (it was always zero) when running the DATA
3258 ACL and the local_scan() function.
3260 19. For the "mail" command in an Exim filter, no default was being set for
3261 the once_repeat time, causing a random time value to be used if "once" was
3262 specified. (If the value happened to be <= 0, no repeat happened.) The
3263 default is now 0s, meaning "never repeat". The "vacation" command was OK
3264 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3265 (I found it when inspecting the code).
3267 20. There is now an overall timeout for performing a callout verification. It
3268 defaults to 4 times the callout timeout, which applies to individual SMTP
3269 commands during the callout. The overall timeout applies when there is more
3270 than one host that can be tried. The timeout is checked before trying the
3271 next host. This prevents very long delays if there are a large number of
3272 hosts and all are timing out (e.g. when the network connections are timing
3273 out). The value of the overall timeout can be changed by specifying an
3274 additional sub-option for "callout", called "maxwait". For example:
3276 verify = sender/callout=5s,maxwait=20s
3278 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3279 to the end before writing, but this should make it even safer).
3281 22. Exim was forgetting that it had advertised PIPELINING for the second and
3282 subsequent messages on an SMTP connection. It was also not resetting its
3283 memory on STARTTLS and an internal HELO.
3285 23. When Exim logs an SMTP synchronization error within a session, it now
3286 records whether PIPELINING has been advertised or not.
3288 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3289 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3290 rather than long int.
3292 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3294 26. Added the never_mail option to autoreply.
3300 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3301 crash if the getsockname() call failed; this can happen if a connection is
3302 closed very soon after it is established. The problem was simply in the
3303 order in which certain operations were done, causing Exim to try to write
3304 to the SMTP stream before it had set up the file descriptor. The bug has
3305 been fixed by making things happen in the correct order.
3311 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3312 before the connection was closed, thus losing the rejection response.
3314 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3315 some early Solaris releases, but causes trouble in current releases where
3316 socklen_t is defined.
3318 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3321 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3324 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3325 that releasing the top of it at the end releases what was used for sub-
3326 expansions (unless the block got too big). However, discard this block if
3327 the first thing is a variable or header, so that we can use its block when
3328 it is dynamic (useful for very large $message_headers, for example).
3330 6. Lookups now cache *every* query, not just the most recent. A new, separate
3331 store pool is used for this. It can be recovered when all lookup caches are
3332 flushed. Lookups now release memory at the end of their result strings.
3333 This has involved some general refactoring of the lookup sources.
3335 7. Some code has been added to the store_xxx() functions to reduce the amount
3336 of flapping under certain conditions.
3338 8. log_incoming_interface used to affect only the <= reception log lines. Now
3339 it causes the local interface and port to be added to several more SMTP log
3340 lines, for example "SMTP connection from", and rejection lines.
3342 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3344 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3346 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3347 could be overwritten at the end of the current message (or the start of a
3348 new message if it was set in a HELO ACL). The value is now preserved for
3349 the duration of the SMTP connection.
3351 12. If a transport had a headers_rewrite setting, and a matching header line
3352 contained an unqualified address, that address was qualified, even if it
3353 did not match any rewriting rules. The underlying bug was that the values
3354 of the flags that permit the existence of unqualified sender and recipient
3355 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3356 non-local messages, and by -bnq for local messages) were not being
3357 preserved with the message after it was received.
3359 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3360 "next input=" as part of the text comprising the host identity instead of
3361 the correct text. The code was using the same buffer for two different
3362 strings. However, depending on which order the printing function evaluated
3363 its arguments, the bug did not always show up. Under Linux, for example, my
3364 test suite worked just fine.
3366 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3367 doesn't work with some older versions of Perl. It has been changed to "my",
3368 which in any case is probably the better facility to use.
3370 15. A really picky compiler found some instances of statements for creating
3371 error messages that either had too many or two few arguments for the format
3374 16. The size of the buffer for calls to the DNS resolver has been increased
3375 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3376 for addresses that have a lot of PTR records. This alleviates a problem; it
3377 does not fully solve it.
3379 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3380 buffer now truncates the list and logs the incident, which is the same
3381 action as happens when Exim is looking up a host name and its aliases.
3382 Previously in this situation something unpredictable would happen;
3383 sometimes it was "internal error: store_reset failed".
3385 18. If a server dropped the connection unexpectedly when an Exim client was
3386 using GnuTLS and trying to read a response, the client delivery process
3387 crashed while trying to generate an error log message.
3389 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3390 string, for example:
3392 warn message = H1: something\nH2: something
3394 the text was added as a single header line from Exim's point of view
3395 though it ended up OK in the delivered message. However, searching for the
3396 second and subsequent header lines using $h_h2: did not work. This has been
3397 fixed. Similarly, if a system filter added multiple headers in this way,
3398 the routers could not see them.
3400 20. Expanded the error message when iplsearch is called with an invalid key to
3401 suggest using net-iplsearch in a host list.
3403 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3404 ACLs are no longer actually imposed (and a message to that effect is
3407 22. If a "gecos" field in a passwd entry contained escaped characters, in
3408 particular, if it contained a \" sequence, Exim got it wrong when building
3409 a From: or a Sender: header from that name. A second bug also caused
3410 incorrect handling when an unquoted " was present following a character
3411 that needed quoting.
3413 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3414 was not being matched caselessly.
3416 24. Arranged for all hyphens in the exim.8 source to be escaped with
3419 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3420 properly. Recipient callout cache records were still being keyed to include
3421 the sender, even when use_sender was set false. This led to far more
3422 callouts that were necessary. The sender is no longer included in the key
3423 when use_sender is false.
3425 26. Added "control = submission" modifier to ACLs.
3427 27. Added the ${base62d: operator to decode base 62 numbers.
3429 28. dnsdb lookups can now access SRV records.
3431 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3432 the configuration file.
3434 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3435 (-v) mode. This makes the output for a verbose queue run more intelligible.
3437 31. Added a use_postmaster feature to recipient callouts.
3439 32. Added the $body_zerocount variable, containing the number of binary zero
3440 bytes in the message body.
3442 33. The time of last modification of the "new" subdirectory is now used as the
3443 "mailbox time last read" when there is a quota error for a maildir
3446 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3448 35. Added +ignore_unknown as a special item in host lists.
3450 36. Code for decoding IPv6 addresses in host lists is now included, even if
3451 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3452 address was recognized as an IP address, but was then not correctly decoded
3453 into binary, causing unexpected and incorrect effects when compared with
3460 1. Very minor rewording of debugging text in manualroute to say "list of
3461 hosts" instead of "hostlist".
3463 2. If verify=header_syntax was set, and a header line with an unqualified
3464 address (no domain) and a large number of spaces between the end of the
3465 name and the colon was received, the reception process suffered a buffer
3466 overflow, and (when I tested it) crashed. This was caused by some obsolete
3467 code that should have been removed. The fix is to remove it!
3469 3. When running in the test harness, delay a bit after writing a bounce
3470 message to get a bit more predictability in the log output.
3472 4. Added a call to search_tidyup() just before forking a reception process. In
3473 theory, someone could use a lookup in the expansion of smtp_accept_max_
3474 per_host which, without the tidyup, could leave open a database connection.
3476 5. Added the variables $recipient_data and $sender_data which get set from a
3477 lookup success in an ACL "recipients" or "senders" condition, or a router
3478 "senders" option, similar to $domain_data and $local_part_data.
3480 6. Moved the writing of debug_print from before to after the "senders" test
3483 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3484 problems for message scanning, either using a data ACL, or using
3485 local_scan() because the Received: header was not generated till after they
3486 were called (in order to set the time as the time of reception completion).
3487 I have revised the way this works. The header is now generated after the
3488 body is received, but before the ACL or local_scan() are called. After they
3489 are run, the timestamp in the header is updated.
3495 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3496 before starting a queue runner without re-exec. This happened only when
3497 deliver_drop_privilege was set or when the Exim user was set to root. The
3498 effect of the bug was that timeouts during subsequent deliveries caused
3499 crashes instead of being properly handled. The handler is now left at its
3500 default (and expected) setting.
3502 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3503 message, again when deliver_drop_privilege is set or Exim is run as root.
3504 The bug described in (1) was not present in this case, but the tidying up
3505 of the other signals was missing. I have made the two cases consistent.
3507 3. The ignore_target_hosts setting on a manualroute router was being ignored
3508 for hosts that were looked up using the /MX notation.
3510 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3513 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3514 operated on the sender address. After changing the $sender_address to <>
3515 for the sender address verify, Exim was re-instated it as the original
3516 (before rewriting) address, but remembering that it had rewritten it, so it
3517 wasn't rewriting it again. This bug also had the effect of breaking the
3518 sender address verification caching when the sender address was rewritten.
3520 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3521 This has been changed so that if the ip literal address matches
3522 ignore_target_hosts, the router declines.
3524 7. Added expansion conditions match_domain, match_address, and match_local_
3525 part (NOT match_host).
3527 8. The placeholder for the Received: header didn't have a length field set.
3529 9. Added code to Exim itself and to exim_lock to test for a specific race
3530 condition that could lead to file corruption when using MBX delivery. The
3531 issue is with the lockfile that is created in /tmp. If this file is removed
3532 after a process has opened it but before that process has acquired a lock,
3533 there is the potential for a second process to recreate the file and also
3534 acquire a lock. This could lead to two Exim processes writing to the file
3535 at the same time. The added code performs the same test as UW imapd; it
3536 checks after acquiring the lock that its file descriptor still refers to
3537 the same named file.
3539 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3540 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3543 11. Added the smtp_active_hostname option. If used, this will typically be made
3544 to depend on the incoming interface address. Because $interface_address is
3545 not set up until the daemon has forked a reception process, error responses
3546 that can happen earlier (such as "too many connections") no longer contain
3549 12. If an expansion in a condition on a "warn" statement fails because a lookup
3550 defers, the "warn" statement is abandoned, and the next ACL statement is
3551 processed. Previously this caused the whole ACL to be aborted.
3553 13. Added the iplsearch lookup type.
3555 14. Added ident_timeout as a log selector.
3557 15. Added tls_certificate_verified as a log selector.
3559 16. Added a global option tls_require_ciphers (compare the smtp transport
3560 option of the same name). This controls incoming TLS connections.
3562 17. I finally figured out how to make tls_require_ciphers do a similar thing
3563 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3564 before starting the TLS session.
3566 18. Tabs are now shown as \t in -bP output.
3568 19. If the log selector return_path_on_delivery was set, Exim crashed when
3569 bouncing a message because it had too many Received: header lines.
3571 20. If two routers both had headers_remove settings, and the first one included
3572 a superfluous trailing colon, the final name in the first list and the
3573 first name in the second list were incorrectly joined into one item (with a
3574 colon in the middle).
3580 1. Added -C and -D options to the exinext utility, mainly to make it easier
3581 to include in the automated testing, but these could be helpful when
3582 multiple configurations are in use.
3584 2. The exinext utility was not formatting the output nicely when there was
3585 an alternate port involved in the retry record key, nor when there was a
3586 message id as well (for retries that were specific to a specific message
3587 and a specific host). It was also confused by IPv6 addresses, because of
3588 the additional colons they contain. I have fixed the IPv4 problem, and
3589 patched it up to do a reasonable job for IPv6.
3591 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3592 delivery, the log line now contains "pipelined" if PIPELINING was used.
3594 4. An SMTP transport process used to panic and die if the bind() call to set
3595 an explicit outgoing interface failed. This has been changed; it is now
3596 treated in the same way as a connect() failure.
3598 5. A reference to $sender_host_name in the part of a conditional expansion
3599 that was being skipped was still causing a DNS lookup. This no longer
3602 6. The def: expansion condition was not recognizing references to header lines
3603 that used bh_ and bheader_.
3605 7. Added the _cache feature to named lists.
3607 8. The code for checking quota_filecount in the appendfile transport was
3608 allowing one more file than it should have been.
3610 9. For compatibility with Sendmail, the command line option
3618 and sets the incoming protocol and host name (for trusted callers). The
3619 host name and its colon can be omitted when only the protocol is to be set.
3620 Note the Exim already has two private options, -pd and -ps, that refer to
3621 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3622 "s", but I don't think that's a major issue.
3624 10. A number of refactoring changes to the code, none of which should affect
3627 (a) The number of logging options was getting close to filling up the
3628 32-bit word that was used as a bit map. I have split them into two classes:
3629 those that are passed in the argument to log_write(), and those that are
3630 only ever tested independently outside of that function. These are now in
3631 separate 32-bit words, so there is plenty of room for expansion again.
3632 There is no change in the user interface or the logging behaviour.
3634 (b) When building, for example, log lines, the code previously used a
3635 macro that called string_cat() twice, in order to add two strings. This is
3636 not really sufficiently general. Furthermore, there was one instance where
3637 it was actually wrong because one of the argument was used twice, and in
3638 one call a function was used. (As it happened, calling the function twice
3639 did not affect the overall behaviour.) The macro has been replaced by a
3640 function that can join an arbitrary number of extra strings onto a growing
3643 (c) The code for expansion conditions now uses a table and a binary chop
3644 instead of a serial search (which was left over from when there were very
3645 few conditions). Also, it now recognizes conditions like "pam" even when
3646 the relevant support is not compiled in: a suitably worded error message is
3647 given if an attempt is made to use such a condition.
3649 11. Added ${time_interval:xxxxx}.
3651 12. A bug was causing one of the ddress fields not to be passed back correctly
3652 from remote delivery subprocesses. The field in question was not being
3653 subsequently used, so this caused to problems in practice.
3655 13. Added new log selectors queue_time and deliver_time.
3657 14. Might have fixed a bug in maildirsizefile handling that threw up
3658 "unexpected character" debug warnings, and recalculated the data
3659 unnecessarily. In any case, I expanded the warning message to give more
3662 15. Added the message "Restricted characters in address" to the statements in
3663 the default ACL that block characters like @ and % in local parts.
3665 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3666 Three changes have been made:
3668 (a) There was a serious bug; a negative response to MAIL caused the whole
3669 recipient domain to be cached as invalid, thereby blocking all messages
3670 to all local parts at the same domain, from all senders. This bug has
3671 been fixed. The domain is no longer cached after a negative response to
3672 MAIL if the sender used is not empty.
3674 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3677 (c) A new callout option, "use_sender" has been added for people who want
3678 the modified behaviour.
3684 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3687 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3688 indeed breaks things for older releases.
3690 3. Added additional logging to the case where there is a problem reading data
3691 from a filter that is running in a subprocess using a pipe, in order to
3692 try to track down a specific problem.
3694 4. Testing facility fudge: when running in the test harness and attempting
3695 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3696 getting "No route to host". Convert this to a timeout.
3698 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3701 6. Some OS don't have socklen_t but use size_t instead. This affects the
3702 fifth argument of getsockopt() amongst other things. This is now
3703 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3704 can be set for individual OS. I have set it for SunOS5, OSF1, and
3705 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3706 some earlier ones do not.
3708 7. Change 4.30/15 was not doing the test caselessly.
3710 8. The standard form for an IPv6 address literal was being rejected by address
3711 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3712 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3713 this, as well as the form without the "IPv6" on the front (but only when
3714 address literals are enabled, of course).
3716 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3718 10. Exim crashed if a message with an empty sender address specified by -f
3719 encountered a router with an errors_to setting. This could be provoked only
3720 by a command such as
3724 where an empty string was supplied; "<>" did not hit this bug.
3726 11. Installed PCRE release 4.5.
3728 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3729 remained set. It is now erased.
3731 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3732 times from message ids (which are base 36 rather than the normal 62).
3734 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3735 were being counted as actual protocol errors, and logged if the log
3736 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3737 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3738 and DATA following a set of rejected RCPTs do not count as protocol errors.
3739 In other words, Exim assumes they were pipelined, though this may not
3740 actually be the case. Of course, in all cases the client gets an
3741 appropriate error code.
3743 15. If a lookup fails in an ACL condition, a message about the failure may
3744 be available; it is used if testing the ACL cannot continue, because most
3745 such messages specify what the cause of the deferral is. However, some
3746 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3747 that caused an old message to be retained and used if a later statement
3748 caused a defer, replacing the real cause of the deferral.
3750 16. If an IP address had so many PTR records that the DNS lookup buffer
3751 was not large enough to hold them, Exim could crash while trying to process
3752 the truncated data. It now detects and logs this case.
3754 17. Further to 4.21/58, another change has been made: if (and only if) the
3755 first line of a message (the first header line) ends with CRLF, a bare LF
3756 in a subsequent header line has a space inserted after it, so as not to
3757 terminate the header.
3759 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3760 unnecessarily, used atoi() instead of strtol(), and didn't check the
3761 termination when getting file sizes from file names by regex.
3763 19. Completely re-implemented the support for maildirsize files, in the light
3764 of a number of problems with the previous contributed implementation
3765 (4.30/29). In particular:
3767 . If the quota is zero, the maildirsize file is maintained, but no quota is
3770 . If the maildir directory does not exist, it is created before any attempt
3771 to write a maildirsize file.
3773 . The quota value in the file is just a cache; if the quota is changed in
3774 the transport, the new value overrides.
3776 . A regular expression is available for excluding directories from the
3779 20. The autoreply transport checks the characters in options that define the
3780 message's headers; it allows continued headers, but it was checking with
3781 isspace() after an embedded newline instead of explicitly looking for a
3784 21. If all the "regular" hosts to which an address was routed had passed their
3785 expiry times, and had not reached their retry times, the address was
3786 bounced, even if fallback hosts were defined. Now Exim should go on to try
3789 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3790 equivalent code in the SMTP transport. Some hosts send humungous responses
3791 to HELO/EHLO, more than 1024 it seems.
3793 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3794 gives compiler warnings in some environments. I've now done it "properly",
3797 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3798 (because of problems with the built-in one) was declared to return uschar *
3799 instead of char *, causing compiler failure.
3801 25. Fixed a file descriptor leak when processing alias/forward files.
3803 26. Fixed a minor format string issue in dbfn.c.
3805 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3807 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3808 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3811 29. When a sender address is verified, it is cached, to save repeating the test
3812 when there is more than one recipient in a message. However, when the
3813 verification involves a callout, it is possible for different callout
3814 options to be set for different recipients. It is too complicated to keep
3815 track of this in the cache, so now Exim always runs a verification when a
3816 callout is required, relying on the callout cache for the optimization.
3817 The overhead is duplication of the address routing, but this should not be
3820 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3821 to be verified with callout=postmaster, and the main callout worked but the
3822 postmaster check failed, the verification correctly failed. However, if a
3823 subsequent RCPT command asked for sender verification *without* the
3824 postmaster check, incorrect caching caused this verification also to fail,
3827 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3828 it was not caching the DNS options (qualify_single, search_parents) that
3829 were used when the lookup failed. A subsequent lookup with different
3830 options therefore always gave the same answer, though there were cases
3831 where it should not have. (Example: a "domains = !$mx_any" option on a
3832 dnslookup router: the "domains" option is always processed without any
3833 widening, but the router might have qualify_single set.) Now Exim uses the
3834 cached value only when the same options are set.
3836 32. Added John Jetmore's "exipick" utility to the distribution.
3838 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3839 than a timeout (e.g. a certificate is required, and is not provided), an
3840 Exim server now closes the connection immediately. Previously it waited for
3841 the client to close - but if the client is SSL, it seems that they each
3842 wait for each other, leading to a delay before one of them times out.
3844 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3845 maintained 0.8.x compatibility because I don't think many are using it, and
3846 it is clearly obsolete.
3848 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3851 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3852 client certificate was expired. A simple patch fixes this, though I don't
3853 understand the full logic of why the verify callback is called multiple
3856 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3857 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3858 which causes problems with some clients (such as the Certicom SSL Plus
3859 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3860 disables the coutermeasure allowing Eudora to connect."
3862 38. Exim was not checking that a write() to a log file succeeded. This could
3863 lead to Bad Things if a log got too big, in particular if it hit a file
3864 size limit. Exim now panics and dies if it cannot write to a log file, just
3865 as it does if it cannot open a log file.
3867 39. Modified OS/Makefile-Linux so that it now contains
3869 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3871 The two -D definitions ensure that Exim is compiled with large file
3872 support, which makes it possible to handle log files that are bigger than
3875 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3876 instance) a domain was checked against a named list that involved a lookup,
3877 causing $domain_data to be set, then another domain was checked against the
3878 same list, then the first domain was re-checked, the value of $domain_data
3879 after the final check could be wrong. In particular, if the second check
3880 failed, it could be set empty. This bug probably also applied to
3883 41. The strip_trailing_dot option was not being applied to the address given
3884 with the -f command-line option.
3886 42. The code for reading a message's header from the spool was incrementing
3887 $received_count, but never initializing it. This meant that the value was
3888 incorrect (doubled) while delivering a message in the same process in which
3889 it was received. In the most common configuration of Exim, this never
3890 happens - a fresh exec is done - but it can happen when
3891 deliver_drop_privilege is set.
3893 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3894 it now includes up to 150 characters of the unexpected data in the log
3897 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3898 and building data strings. The size of both of these buffers was 10 000
3899 bytes - far larger than anybody would *ever* want, thought I. Needless to
3900 say, somebody hit the limit. I have increased the maximum line length to
3901 20 000 and the maximum data length of concatenated lines to 100 000. I have
3902 also fixed two bugs, because there was no checking on these buffers. Tsk,
3903 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3904 buffer is too small.
3906 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3907 lsearch lookups. Now it does.
3909 46. When parsing a route_list item in a manualroute router, a fixed-length
3910 buffer was used for the list of hosts. I made this 1024 bytes long,
3911 thinking that nobody would ever have a list of hosts that long. Wrong.
3912 Somebody had a whole pile of complicated expansion conditions, and the
3913 string was silently truncated, leading to an expansion error. It turns out
3914 that it is easier to change to an unlimited length (owing to other changes
3915 that have happened since this code was originally written) than to build
3916 structure for giving a limitation error. The length of the item that
3917 expands into the list of hosts is now unlimited.
3919 47. The lsearch lookup could not handle data where the length of text line was
3920 more than 4095 characters. Such lines were truncated, leading to shortened
3921 data being returned. It should now handle lines of any length.
3923 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3924 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3927 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3928 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3929 no longer happens in comment lines. A list of replacements is now placed
3930 at the head of all of the source files, except those whose only change is
3931 to replace PERL_COMMAND in the very first #! line.
3933 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3934 messages on the queue, with a bottom-up merge sort, using code contributed
3935 by Michael Haardt. This should make operations like -bp somewhat faster on
3936 large queues. It won't affect queue runners, except when queue_run_in_order
3939 51. Installed eximstats 1.31 in the distribution.
3941 52. Added support for SRV lookups to the dnslookup router.
3943 53. If an ACL referred to $message_body or $message_body_end, the value was not
3944 reset for any messages that followed in the same SMTP session.
3946 54. The store-handling optimization for building very long strings was not
3947 differentiating between the different store pools. I don't think this
3948 actually made any difference in practice, but I've tidied it.
3950 55. While running the routers to verify a sender address, $sender_address
3951 was still set to the sender address. This is wrong, because when routing to
3952 send a bounce to the sender, it would be empty. Therefore, I have changed
3953 it so that, while verifying a sender address, $sender_address is set to <>.
3954 (There is no change to what happens when verifying a recipient address.)
3956 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3957 target A or AAAA records (if not already returned) without resetting the
3958 qualify_single or search_parents options of the DNS resolver. These are
3959 inappropriate in this case because the targets of MX and SRV records must
3960 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3961 target that, when qualified, matched something in the local domain. These
3962 two options are now turned off when doing these lookups.
3964 57. It seems that at least some releases of Reiserfs (which does not have the
3965 concept of a fixed number of inodes) returns zero and not -1 for the
3966 number of available inodes. This interacted badly with check_spool_inodes,
3967 which assumed that -1 was the "no such thing" setting. What I have done is
3968 to check that the total number of inodes is greater than zero before doing
3969 the test of how many are available.
3971 58. When a "warn" ACL statement has a log_message modifier, the message is
3972 remembered, and not repeated. This is to avoid a lot of repetition when a
3973 message has many recipients that cause the same warning to be written.
3974 Howewer, Exim was preserving the list of already written lines for an
3975 entire SMTP session, which doesn't seem right. The memory is now reset if a
3976 new message is started.
3978 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3979 debugging output unless log_rewrite was also set.
3981 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3982 of (int)(handle) when we know that handle contains (void *)(-1).
3984 61. The Exim daemon panic-logs an error return when it closes the incoming
3985 connection. However "connection reset by peer" seems to be common, and
3986 isn't really an error worthy of noting specially, so that particular error
3989 62. When Exim is trying to find all the local interfaces, it used to panic and
3990 die if the ioctl to get the interface flags failed. However, it seems that
3991 on at least one OS (Solaris 9) it is possible to have an interface that is
3992 included in the list of interfaces, but for which you get a failure error
3993 for this call. This happens when the interface is not "plumbed" into a
3994 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3995 failure of the "get flags" call assumes that the interface is down.
3997 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3998 makes life easier for people who are doing arithmetic on fields extracted
3999 from dates, where you often get leading zeros that should not be
4000 interpreted as octal.
4002 64. Added qualify_domain to the redirect router, to override the global
4005 65. If a pathologically long header line contained very many addresses (the
4006 report of this problem mentioned 10 000) and each of them was rewritten,
4007 Exim could use up a very large amount of memory. (It kept on making new
4008 copies of the header line as it rewrote, and never released the old ones.)
4009 At the expense of a bit more processing, the header rewriting function has
4010 been changed so that it no longer eats memory in this way.
4012 66. The generation of the Received: header has been moved from the time that a
4013 message starts to be received, to the time that it finishes. The timestamp
4014 in the Received: header should now be very close to that of the <= log
4015 line. There are two side-effects of this change:
4017 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4018 logged header lines no longer include the local Received: line, because
4019 it has not yet been created. The same applies to a copy of the message
4020 that is returned to a non-SMTP sender when a message is rejected.
4022 (b) When a filter file is tested using -bf, no additional Received: header
4023 is added to the test message. After some thought, I decided that this
4026 This change does not affect the value of $received_for. It is still set
4027 after address rewriting, but before local_scan() is called.
4029 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4031 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4032 gave an unhelpful panic error message, and a defer error. I have managed to
4033 change this behaviour so that it now rejects any supplied certificate,
4034 which seems right, as the list of acceptable certificates is empty.
4036 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4037 gave an unhelpful defer error. I have not managed to make this reject any
4038 supplied certificates, but the error message it gives is "no certificate
4039 supplied", which is not helpful.
4041 70. exigrep's output now also includes lines that are not associated with any
4042 message, but which match the given pattern. Implemented by a patch from
4043 Martin Sluka, which also tidied up the Perl a bit.
4045 71. Recipient callout verification, like sender verification, was using <> in
4046 the MAIL FROM command. This isn't really the right thing, since the actual
4047 sender may affect whether the remote host accepts the recipient or not. I
4048 have changed it to use the actual sender in the callout; this means that
4049 the cache record is now keyed on a recipient/sender pair, not just the
4050 recipient address. There doesn't seem to be a real danger of callout loops,
4051 since a callout by the remote host to check the sender would use <>.
4052 [SEE ABOVE: changed after hitting problems.]
4054 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4055 temporary errors. However, in the case of such a code being given after
4056 the end of a data transmission (i.e. after ".") Exim was failing to write
4057 a retry record for the message. (Yes, there was some broken host that was
4058 actually sending 8xx at this point.)
4060 73. An unknown lookup type in a host list could cause Exim to panic-die when
4061 the list was checked. (An example that provoked this was putting <; in the
4062 middle of a list instead of at the start.) If this happened during a DATA
4063 ACL check, a -D file could be left lying around. This kind of configuration
4064 error no longer causes Exim to die; instead it causes a defer errror. The
4065 incident is still logged to the main and panic logs.
4067 74. Buglet left over from Exim 3 conversion. The message "too many messages
4068 in one connection" was written to the rejectlog but not the mainlog, except
4069 when address rewriting (yes!) was being logged.
4071 75. Added write_rejectlog option.
4073 76. When a system filter was run not as root (that is, when system_filter_user
4074 was set), the values of the $n variables were not being returned to the
4075 main process; thus, they were not subsequently available in the $sn
4078 77. Added +return_path_on_delivery log selector.
4080 78. A connection timeout was being treated differently from recipients deferred
4081 when testing hosts_max_try with a message that was older than the host's
4082 retry timeout. (The host should not be counted, thus allowing all hosts to
4083 be tried at least once before bouncing.) This may have been the cause of an
4084 occasionally reported bug whereby a message would remain on the queue
4085 longer than the retry timeout, but would be bounced if a delivery was
4086 forced. I say "may" because I never totally pinned down the problem;
4087 setting up timeout/retry tests is difficult. See also the next item.
4089 79. The ultimate address timeout was not being applied to errors that involved
4090 a combination of host plus message (for example, a timeout on a MAIL
4091 command). When an address resolved to a number of possible hosts, and they
4092 were not all tried for each delivery (e.g. because of hosts_max_try), a
4093 message could remain on the queue longer than the retry timeout.
4095 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4098 81. Fixed an obscure SMTP outgoing bug which required at least the following
4099 conditions: (a) there was another message waiting for the same server;
4100 (b) the server returned 5xx to all RCPT commands in the first message so
4101 that the message was not completed; (c) the server dropped the connection
4102 or gave a negative response to the RSET that Exim sends to abort the
4103 transaction. The observed case was a dropped connection after DATA that had
4104 been sent in pipelining mode. That is, the server had advertised PIPELINING
4105 but was not implementing it correctly. The effect of the bug was incorrect
4106 behaviour, such as trying another host, and this could lead to a crash.
4112 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4113 and daemon.c were passed as pointers to ints; they should have been
4114 pointers to socklen_t variables (which are typically unsigned ints).
4116 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4119 3. Fixed a really odd bug that affected only the testing scheme; patching a
4120 certain fixed string in the binary changed the value of another string that
4121 happened to be identical to the end of the original first string.
4123 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4124 name", it returns that address as the IP address. On some operating
4125 systems (e.g. Solaris), it also passes back the IP address string as the
4126 "host name". However, on others (e.g. Linux), it passes back an empty
4127 string. Exim wasn't checking for this, and was changing the host name to an
4128 empty string, assuming it had been canonicized.
4130 5. Although rare, it is permitted to have more than one PTR record for a given
4131 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4132 all the names associated with an address, because they do in Solaris.
4133 However, it seems that they do not in Linux for data that comes from the
4134 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4135 I found this out when I moved to a new Linux workstation and tried to run
4136 the Exim test suite.
4138 To get round this problem I have changed the code so that it now does its
4139 own call to the DNS to look up PTR records when searching for a host name.
4140 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4141 addresses that are only in /etc/hosts are still found.
4143 This behaviour is, however, controlled by an option called host_lookup_
4144 order, which defaults to "bydns:byaddr". If people want to use the other
4145 order, or indeed, just use one or the other means of lookup, they can
4146 specify it in this variable.
4148 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4149 some operating systems, this comes back from gethostbyaddr() as an empty
4150 string, and this is what Exim used to test for. However, it seems that in
4151 other systems, "." is yielded. Exim now tests for this case too.
4153 7. The values of check_spool_space and check_log_space are now held internally
4154 as a number of kilobytes instead of an absolute number of bytes. If a
4155 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4156 kilobyte. This means that much larger values can be stored.
4158 8. Exim monitor: an attempt to get the action menu when not actually pointing
4159 at a message produces an empty menu entitled "No message selected". This
4160 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4161 no entries in it ("Shell widget menu has zero width and/or height"). So I
4162 have added a single, blank menu entry in this case.
4164 9. Added ${quote_local_part.
4166 10. MIME decoding is now applied to the contents of Subject: header lines when
4169 11. Now that a reference to $sender_host_address automatically causes a reverse
4170 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4171 host lookup before query-style lookups in lists that might use this
4172 variable. This has therefore been abolished, and the "net-" prefix is no
4173 longer necessary for query-style lookups.
4175 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4176 have been a typo for LFLAGS, so it has been changed.
4178 13. The install script calls Exim with "-C /dev/null" in order to find the
4179 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4180 to be output. Howeve, since Exim outputs its version number before the
4181 error, it didn't break the script. It just looked ugly. I fixed this by
4182 always allowing "-C /dev/null" if the caller is root.
4184 14. Ignore overlarge ACL variable number when reading spool file - insurance
4185 against a later release with more variables having written the file.
4187 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4188 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4189 this, as well as the form without the "IPv6" on the front.
4191 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4192 OS/Makefile-Darwin file.
4194 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4195 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4197 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4198 variable. However, some OS use ints for inodes. I've added cast to long int
4199 to get rid of the compiler warning.
4201 19. I had forgotten to lock out "/../" in configuration file names when
4202 ALT_CONFIG_PREFIX was set.
4204 20. Routers used for verification do not need to specify transports. However,
4205 if such a router generated a host list, and callout was configured, Exim
4206 crashed, because it could not find a port number from the (non-existent)
4207 transport. It now assumes port 25 in this circumstance.
4209 21. Added the -t option to exigrep.
4211 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4212 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4213 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4214 them is set, LOOKUP_LSEARCH is forced.
4216 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4218 24. Added sender and host information to the "rejected by local_scan()" log
4219 line; previously there was no indication of these.
4221 25. Added .include_if_exists.
4223 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4224 Linux. It turns out that not all file systems support this. Apparently some
4225 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4226 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4227 not supported on the file descriptor, is now ignored when Exim is trying to
4228 sync a directory. This applies only to Linux.
4230 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4232 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4233 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4236 29. Added support for maildirsize files from supplied patch (modified a bit).
4238 30. The use of :fail: followed by an empty string could lead Exim to respond to
4239 sender verification failures with (e.g.):
4241 550 Verification failed for <xxx>
4242 550 Sender verify failed
4244 where the first response line was missing the '-' that indicates it is not
4245 the final line of the response.
4247 31. The loop for finding the name of the user that called Exim had a hardwired
4248 limit of 10; it now uses the value of finduser_retries, which is used for
4249 all other user lookups.
4251 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4254 33. Exim was neglecting to zero errno before one call of strtol() when
4255 expanding a string and expecting an integer value. On some systems this
4256 resulted in spurious "integer overflow" errors. Also, it was casting the
4257 result into an int without checking.
4259 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4260 did not work. The code looks as if it has *never* worked, though it appears
4261 to have been documented since at least releast 1.62. I have made it work.
4263 35. The "timeout_DNS" error in retry rules, also documented since at least
4264 1.62, also never worked. As it isn't clear exactly what this means, and
4265 clearly it isn't a major issue, I have abolished the feature by treating it
4266 as "timeout", and writing a warning to the main and panic logs.
4268 36. The display of retry rules for -brt wasn't always showing the error code
4271 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4272 timeout_connect_A, timeout_connect_MX.
4274 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4275 to the empty sender.
4277 39. The daemon was not analysing the content of -oX till after it had closed
4278 stderr and disconnected from the controlling terminal. This meant that any
4279 syntax errors were only noted on the panic log, and the return code from
4280 the command was 0. By re-arranging the code a little, I've made the
4281 decoding happen first, so such errors now appear on stderr, and the return
4282 code is 1. However, the actual setting up of the sockets still happens in
4283 the disconnected process, so errors there are still only recorded on the
4286 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4287 connections (as happens on some IP stacks) was logged at start up time as
4288 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4289 it from "IPv6 and IPv4", which means that two separate sockets are being
4292 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4293 says whether AF_INET or AF_INET6 was passed as an argument.
4295 42. Exiwhat output was messed up when time zones were included in log
4298 43. Exiwhat now gives more information about the daemon's listening ports,
4299 and whether -tls-on-connect was used.
4301 44. The "port" option of the smtp transport is now expanded.
4303 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4304 silently ignored. Now an error message is written to the main and panic
4307 46. There's a new ACL modifier called "logwrite" which writes to a log file
4308 as soon as it is encountered.
4310 47. Added $local_user_uid and $local_user_gid at routing time.
4312 48. Exim crashed when trying to verify a sender address that was being
4315 49. Exim was recognizing only a space character after ".include". It now also
4316 recognizes a tab character.
4318 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4319 extracting the relevant information from the specification. The man page no
4320 longer contains scrambled data for the -d option, and I've added a section
4321 at the front about calling Exim under different names.
4323 51. Added "extra_headers" argument to the "mail" command in filter files.
4325 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4328 53. Installed eximstats 1.29.
4330 54. Added transport_filter_timeout as a generic transport option.
4332 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4333 Cc: header lines. This was required by RFC 822, but it not required by RFC
4336 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4337 incoming messages that did not have them. Now it does so only if the
4338 message originates locally, that is, if there is no associated remote host
4339 address. When Resent- header lines are present, this applies to the Resent-
4340 lines rather than the non-Resent- lines.
4342 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4343 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4345 58. Messages for configuration errors now include the name of the main
4346 configuration file - useful now that there may be more than one file in a
4347 list (.included file names were always shown).
4349 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4350 for those rare installations that do not start the daemon as root or run it
4351 setuid root. I've cut out the call to initgroups() if the daemon is not
4354 60. The Exim user and group can now be bound into the binary as text strings
4355 that are looked up at the start of Exim's processing.
4357 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4359 62. Added $mailstore_basename variable.
4361 63. Installed patch to sieve.c from Michael Haardt.
4363 64. When Exim failed to open the panic log after failing to open the main log,
4364 the original message it was trying to log was written to stderr and debug
4365 output, but if they were not available (the usual case in production), it
4366 was lost. Now it is written to syslog before the two lines that record the
4367 failures to open the logs.
4369 65. Users' Exim filters run in subprocesses under the user's uid. It is
4370 possible for a "deliver" command or an alias in a "personal" command to
4371 provoke an address rewrite. If logging of address rewriting is configured,
4372 this fails because the process is not running as root or exim. There may be
4373 a better way of dealing with this, but for the moment (because 4.30 needs
4374 to be released), I have disabled address rewrite logging when running a
4375 filter in a non-root, non-exim process.
4381 1. The buildconfig auxiliary program wasn't quoting the value set for
4382 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4383 not defined. This bug was masked in 4.22 by the effect that was fixed in
4386 2. Some messages that were rejected after a message id was allocated were
4387 shown as "incomplete" by exigrep. It no longer does this for messages that
4388 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4390 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4391 have allow_domain_literals set, the ID did not get logged in the <= line.
4392 Domain literals are now always recognized in Message-ID: header lines.
4394 4. The first argument for a ${extract expansion item is the key name or field
4395 number. Leading and trailing spaces in this item were not being ignored,
4396 causing some misleading effects.
4398 5. When deliver_drop_privilege was set, single queue runner processes started
4399 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4400 same command in the process it spins off) were not dropping privilege.
4402 6. When the daemon running as "exim" started a queue runner, it always
4403 re-executed Exim in the spun-off process. This is a waste of effort when
4404 deliver_drop_privilege is set. The new process now just calls the
4405 queue-runner function directly.
4411 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4414 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4415 ignored. Though the use of -oP was forcing the writing of a pid file, it
4416 was always written to the default place.
4418 3. If the message "no IP address found for host xxxx" is generated during
4419 incoming verification, it is now followed by identification of the incoming
4420 connection (so you can more easily find what provoked it).
4422 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4424 5. Added some features to "harden" Exim a bit more against certain attacks:
4426 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4427 be put in Local/Makefile. This is like the never_users runtime option,
4428 but it cannot be overridden. The default setting is "root".
4430 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4431 prefix string with which any file named in a -C command line option
4434 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4435 is retained for -C and -D only if the caller of Exim is root. Without
4436 it, the exim user may also use -C and -D and retain privilege.
4438 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4439 command line option is disabled.
4441 6. Macro names set by the -D option must start with an upper case letter, just
4442 like macro names defined in the configuration file.
4444 7. Added "dereference=" facility to LDAP.
4446 8. Two instances of the typo "uknown" in the source files are fixed.
4448 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4449 the Configure-Makefile script screwed up while processing it.
4451 10. Incorporated PCRE 4.4.
4453 11. The SMTP synchronization check was not operating right at the start of an
4454 SMTP session. For example, it could not catch a HELO sent before the client
4455 waited for the greeting. There is now a check for outstanding input at the
4456 point when the greeting is written. Because of the duplex, asynchronous
4457 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4458 way, but not yet received, when the check is performed.
4460 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4461 on TCP/IP sockets, because this apparently causes some broken clients to
4464 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4465 unchanged) from the Cygwin maintainer.
4467 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4468 was supported instead of testing for mbx. Effectively a typo.
4470 15. The spa authenticator server code was not checking that the input it
4471 received was valid base64.
4473 16. The debug output line for the "set" modifier in ACLs was not showing the
4474 name of the variable that was being set.
4476 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4478 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4479 lookup on its own. Something else was needed to trigger the lookup. For
4480 example, a match in host_lookup or the need for a host name in a host list.
4481 Now, if $sender_host_name is referenced and the host name has not yet been
4482 looked up, a lookup is performed. If the lookup fails, the variable remains
4483 empty, and $host_lookup_failed is set to "1".
4485 19. Added "eqi" as a case-independent comparison operator.
4487 20. The saslauthd authentication condition could segfault if neither service
4488 nor realm was specified.
4490 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4491 error message that was logged was misleading, and incoming SMTP
4492 connections were dropped. The message is now more accurate, and temporary
4493 errors are given to SMTP connections.
4495 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4496 (except RSET) with a 421 error, until QUIT is received. However, it was
4497 failing to send a response to QUIT.
4499 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4500 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4501 ACL is now run after the helo_try_verify_hosts code.
4503 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4504 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4505 case-independent, so other case variants are also recognized). Apparently
4506 some systems use these upper case variants.
4508 25. If more than two messages were waiting for the same host, and a transport
4509 filter was specified for the transport, Exim sent two messages over the
4510 same TCP/IP connection, and then failed with "socket operation on non-
4511 socket" when it tried to send the third.
4513 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4515 27. The extern definition of crypt16() in expand.c was not being excluded when
4516 the OS had its own crypt16() function.
4518 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4519 as a preferred synonym for return_size_limit, both as an option and as an
4522 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4524 30. Changed the default configuration ACL to relax the local part checking rule
4525 for addresses that are not in any local domains. For these addresses,
4526 slashes and pipe symbols are allowed within local parts, but the sequence
4527 /../ is explicitly forbidden.
4529 31. SPA server authentication was not clearing the challenge buffer before
4532 32. log_message in a "warn" ACL statement was writing to the reject log as
4533 well as to the main log, which contradicts the documentation and doesn't
4534 seem right (because no rejection is happening). So I have stopped it.
4536 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4537 However, I am unable to do any testing of this.
4539 34. Fixed an infelicity in the appendfile transport. When checking directories
4540 for a mailbox, to see if any needed to be created, it was accidentally
4541 using path names with one or more superfluous leading slashes; tracing
4542 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4544 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4545 contents are added to the log line that is written for every discarded
4546 recipient. (Previously a log_message setting was ignored.)
4548 36. The ${quote: operator now quotes the string if it is empty.
4550 37. The install script runs exim in order to find its version number. If for
4551 some reason other than non-existence or emptiness, which it checks, it
4552 could not run './exim', it was installing it with an empty version number,
4553 i.e. as "exim-". This error state is now caught, and the installation is
4556 38. An argument was missing from the function that creates an error message
4557 when Exim fails to connect to the socket for saslauthd authentication.
4558 This could cause Exim to crash, or give a corrupted message.
4560 39. Added isip, isip4, and isip6 to ${if conditions.
4562 40. The ACL variables $acl_xx are now saved with the message, and can be
4563 accessed later in routers, transports, and filters.
4565 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4566 strings in the file are not string-expanded.
4568 42. If a MAIL command specified a SIZE value that was too large to fit into an
4569 int variable, the check against message_size_limit failed. Such values are
4570 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4571 day this will have to be increased, but I don't think I want to be around
4572 when emails are that large.
4579 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4580 iconv() is not standard in FreeBSD.
4582 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4583 IPv6 enabled. The observed symptom was a segmentation fault on return from
4584 the function os_common_find_running_interfaces() in src/os.c.
4586 3. In the check_special_case() function in daemon.c I had used "errno" as an
4587 argument name, which causes warnings on some systems. This was basically a
4588 typo, since it was named "eno" in the comments!
4590 4. The code that waits for the clock to tick (at a resolution of some fraction
4591 of a second) so as to ensure message-id uniqueness was always waiting for
4592 at least one whole tick, when it could have waited for less. [This is
4593 almost certainly not relevant at current processor speeds, where it is
4594 unlikely to ever wait at all. But we try to future-proof.]
4596 5. The function that sleeps for a time interval that includes fractions of a
4597 second contained a race. It did not block SIGALRM between setting the
4598 timer, and suspending (a couple of lines later). If the interval was short
4599 and the sigsuspend() was delayed until after it had expired, the suspension
4600 never ended. On busy systems this could lead to processes getting stuck for
4603 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4604 process, before it forks any delivery processes. The open lookup caching
4605 mechanism meant that the open file or database connection was passed into
4606 the delivery process. The problem was that delivery processes always tidy
4607 up cached lookup data. This could cause a problem for the next delivery
4608 process started by the queue runner, because the external queue runner
4609 process does not know about the closure. So the next delivery process
4610 still has data in the lookup cache. In the case of a file lookup, there was
4611 no problem because closing a file descriptor in a subprocess doesn't affect
4612 the parent. However, if the lookup was caching a connection to a database,
4613 the connection was closed, and the second delivery process was likely to
4614 see errors such as "PGSQL: query failed: server closed the connection
4615 unexpectedly". The problem has been fixed by closing all cached lookups
4616 in a queue runner before running a delivery process.
4618 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4619 seem to have the "const" qualifier which it has on other OS. I've
4622 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4623 *of the same type* (client or server) with the same public name that an
4624 error should be diagnosed.
4626 9. When Exim looked up a host name for an IP address, but failed to find the
4627 original IP address when looking up the host name (a safety check), it
4628 output the message "<ip address> does not match any IP for NULL", which was
4629 confusing, to say the least. The bug was that the host name should have
4630 appeared instead of "NULL".
4632 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4633 user that is built into the binary, and the -C or -D options is used, root
4634 privilege is dropped before the configuration file is read. In addition,
4635 logging is switched to stderr instead of the normal log files. If the
4636 configuration then re-defines the Exim user, the unprivileged environment
4637 is probably not what is expected, so Exim logs a panic warning message (but
4640 However, if deliver_drop_privilege is set, the unprivileged state may well
4641 be exactly what is intended, so the warning has been cut out in that case,
4642 and Exim is allowed to try to write to its normal log files.
4648 1. smtp_return_error_details was not giving details for temporary sender
4649 or receiver verification errors.
4651 2. Diagnose a configuration error if two authenticators have the same public
4654 3. Exim used not to create the message log file for a message until the first
4655 delivery attempt. This could be confusing when incoming messages were held
4656 for policy or load reasons. The message log file is now created at the time
4657 the message is received, and an initial "Received" line is written to it.
4659 4. The automatically generated man page for command line options had a minor
4660 bug that caused no ill effects; however, a more serious problem was that
4661 the procedure for building the man page automatically didn't always
4662 operate. Consequently, release 4.20 contains an out-of-date version. This
4663 shouldn't happen again.
4665 5. When building Exim with embedded Perl support, the script that builds the
4666 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4667 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4669 6. The freeze_tell option was not being used for messages that were frozen on
4670 arrival, either by an ACL or by local_scan().
4672 7. Added the smtp_incomplete_transaction log selector.
4674 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4675 was accepting AUTH without a new EHLO.
4677 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4678 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4681 10. Logging of TCP/IP connections (when configured) now happens in the main
4682 daemon process instead of the child process, so that the TCP/IP connection
4683 count is more accurate (but it can never be perfect).
4685 11. The use of "drop" in a nested ACL was not being handled correctly in the
4686 outer ACL. Now, if condition failure induced by the nested "drop" causes
4687 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4688 or "require"), the connection is dropped.
4690 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4691 that yield "discard" can now be used with an "accept" or a "discard" verb,
4692 but an error is generated for any others (because I can't see a useful way
4693 to define what should happen).
4695 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4696 are now processed in the same way as lines in the Exim configuration file.
4697 In particular, continuation lines are supported.
4699 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4701 15. Added -ti meaning -t -i.
4703 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4704 domains, and warn by logging if others are found.
4706 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4707 returned from GIFCONF when Exim is trying to find the list of interfaces on
4708 a host. The code in os.c has been modified to copy each ifreq to an aligned
4709 structure in all cases.
4711 Also, in some cases, the returned ifreq's were being copied to a 'struct
4712 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4713 means the last couple of bytes of an IPv6 address could be chopped if the
4714 ifreq contained only a normal sockaddr (14 bytes storage).
4716 18. Named domain lists were not supported in the hosts_treat_as_local option.
4717 An entry such as +xxxx was not recognized, and was treated as a literal
4720 19. Ensure that header lines added by a DATA ACL are included in the reject log
4721 if the ACL subsequently rejects the message.
4723 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4724 MD5 (which is deprecated).
4726 21. When testing a filter file using -bf, Exim was writing a message when it
4727 took the sender from a "From " line in the message, but it was not doing so
4728 when it took $return_path from a Return-Path: header line. It now does.
4730 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4731 with a valid header line field name (a series of printing characters
4732 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4734 23. Changed "disc" in the source to "disk" to conform to the documentation and
4735 the book and for uniformity.
4737 24. Ignore Sendmail's -Ooption=value command line item.
4739 25. When execve() failed while trying to run a command in a pipe transport,
4740 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4741 could be confused with a return value of 69 from the command itself. This
4742 has been changed to 127, the value the shell returns if it is asked to run
4743 a non-existent command. The wording for the related log line suggests a
4744 non-existent command as the problem.
4746 26. If received_header_text expands to an empty string, do not add a Received:
4747 header line to the message. (Well, it adds a token one on the spool, but
4748 marks it "old" so that it doesn't get used or transmitted.)
4750 27. Installed eximstats 1.28 (addition of -nt option).
4752 28. There was no check for failure on the call to getsockname() in the daemon
4753 code. This can fail if there is a shortage of resources on the system, with
4754 ENOMEM, for example. A temporary error is now given on failure.
4756 29. Contrary to the C standard, it seems that in some environments, the
4757 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4758 program. Exim now does this explicitly; it affects the formatting of
4759 timestamps using strftime().
4761 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4762 complaints. I've now initialized all the variables, to avoid this.
4764 32. Header lines added by a system filter were not being "seen" during
4765 transport-time rewrites.
4767 33. The info_callback() function passed to OpenSSL is set up with type void
4768 (*)(SSL *, int, int), as described somewhere. However, when calling the
4769 function (actually a macro) that sets it up, the type void(*)() is
4770 expected. I've put in a cast to prevent warnings from picky compilers.
4772 34. If a DNS black list lookup found a CNAME record, but there were no A
4773 records associated with the domain it pointed at, Exim crashed.
4775 35. If a DNS black list lookup returned more than one A record, Exim ignored
4776 all but the first. It now scans all returned addresses if a particular IP
4777 value is being sought. In this situation, the contents of the
4778 $dnslist_value variable are a list of all the addresses, separated by a
4781 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4782 to accept a host name and all its aliases if the forward lookup for any of
4783 them yielded the IP address of the incoming connection. Now it accepts only
4784 those names whose forward lookup yields the correct IP address. Any other
4785 names are discarded. This closes a loophole whereby a rogue DNS
4786 administrator could create reverse DNS records to break through a
4787 wildcarded host restriction in an ACL.
4789 37. If a user filter or a system filter that ran in a subprocess used any of
4790 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4791 the wrong values were passed to the pipe command ($thisaddress had the
4792 value of $0, $0 had the value of $1, etc). This bug was introduced by
4793 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4795 38. Improved the line breaking for long SMTP error messages from ACLs.
4796 Previously, if there was no break point between 40 and 75 characters, Exim
4797 left the rest of the message alone. Two changes have been made: (a) I've
4798 reduced the minimum length to 35 characters; (b) if it can't find a break
4799 point between 35 and 75 characters, it looks ahead and uses the first one
4800 that it finds. This may give the occasional overlong line, but at least the
4801 remaining text gets split now.
4803 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4804 file descriptors might be low, and that setting 1000 would always raise it.
4805 It turns out that in some environments, the limit is already over 1000 and
4806 that lowering it causes trouble. So now Exim takes care not to decrease it.
4808 40. When delivering a message, the value of $return_path is set to $sender_
4809 address at the start of routing (routers may change the value). By an
4810 oversight, this default was not being set up when an address was tested by
4811 -bt or -bv, which affected the outcome if any router or filter referred to
4814 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4815 log only when "message" or "log_message" are set. However, if one of the
4816 conditions was an address verification, or a call to a nested ACL, the
4817 messages generated by the underlying test were being passed through. This
4818 no longer happens. The underlying message is available in $acl_verify_
4819 message for both "message" and "log_message" expansions, so it can be
4820 passed through if needed.
4822 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4823 new expansion $bh_ to give the encoded byte string without charset
4824 translation. Translation happens only if iconv() is available; HAVE_ICONV
4825 indicates this at build time. HEADERS_CHARSET gives the charset to
4826 translate to; headers_charset can change it in the configuration, and
4827 "headers charset" can change it in an individual filter file.
4829 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4830 that creates RFC 2047 encoded "words" labels them as that charset instead
4831 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4832 expansion operator; (ii) when Exim creates a From: line for a local
4833 message; (iii) when a header line is rewritten to include a "phrase" part.
4835 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4836 buggy, causing it to skip the first lines of messages whose message ID
4837 ended in 'D'. This would not have bitten before Exim release 4.14, because
4838 message IDs were unlikely to end in 'D' before then. The effect was to have
4839 incorrect size information for certain domains.
4841 45. #include "config.h" was missing at the start of the crypt16.c module. This
4842 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4845 46. If there was a timeout during a "random" callout check, Exim treated it as
4846 a failure of the random address, and carried on sending RSET and the real
4847 address. If the delay was just some slowness somewhere, the response to the
4848 original RCPT would be taken as a response to RSET and so on, causing
4849 mayhem of various kinds.
4851 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4852 when I implemented it. It didn't allow for the fact that some option values
4853 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4854 the right test for positive values.
4856 48. Domain names in DNS records are case-independent. Exim always looks them up
4857 in lower case. Some resolvers return domain names in exactly the case they
4858 appear in the zone file, that is, they may contain uppercase letters. Not
4859 all resolvers do this - some return always lower case. Exim was treating a
4860 change of case by a resolver as a change of domain, similar to a widening
4861 of a domain abbreviation. This triggered its re-routing code and so it was
4862 trying to route what was effectively the same domain again. This normally
4863 caused routing to fail (because the router wouldn't handle the domain
4864 twice). Now Exim checks for this case specially, and just changes the
4865 casing of the domain that it ultimately uses when it transmits the message
4868 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4871 50. If a filter generated a file delivery with a non-absolute name (possible if
4872 no home directory exists for the router), the forbid_file option was not
4875 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4876 the existing equality matching.
4878 52. Exim was using ints instead of ino_t variables in some places where it was
4879 dealing with inode numbers.
4881 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4882 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4883 called TMPDIR, and if it finds it is different, it changes its value.
4885 54. The smtp_printf() function is now made available to local_scan() so
4886 additional output lines can be written before returning. There is also an
4887 smtp_fflush() function to enable the detection of a dropped connection.
4888 The variables smtp_input and smtp_batched_input are exported to
4891 55. Changed the default runtime configuration: the message "Unknown user"
4892 has been removed from the ACL, and instead placed on the localuser router,
4893 using the cannot_route_message feature. This means that any verification
4894 failures that generate their own messages won't get overridden. Similarly,
4895 the "Unrouteable address" message that was in the ACL for unverifiable
4896 relay addresses has also been removed.
4898 56. Added hosts_avoid_esmtp to the smtp transport.
4900 57. The exicyclog script was not checking for the esoteric option
4901 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4902 will work only if exicyclog is run under the appropriate euid.
4904 58. Following a discussion on the list, the rules by which Exim recognises line
4905 endings on incoming messages have been changed. The -dropcr and drop_cr
4906 options are now no-ops, retained only for backwards compatibility. The
4907 following line terminators are recognized: LF CRLF CR. However, special
4908 processing applies to CR:
4910 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4911 nor a local message in the state where . is a terminator.
4913 (ii) If a bare CR is encountered in a header line, an extra space is added
4914 after the line terminator so as not to end the header. The reasoning
4915 behind this is that bare CRs in header lines are most likely either
4916 to be mistakes, or people trying to play silly games.
4918 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4919 was being incorrectly given as 18 bytes larger than it should have been.
4920 This is a VOB (very old bug).
4922 60. This may never have affected anything current, but just in case it has:
4923 When the local host is found other than at the start of a list of hosts,
4924 the local host, those with the same MX, and any that follow, are discarded.
4925 When the list in question was part of a longer list of hosts, the following
4926 hosts (not currently being processed) were also being discarded. This no
4927 longer happens. I'm not sure if this situation could ever has previously
4930 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4933 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4934 header. This is something that is recommended in a new Internet Draft, and
4935 is something that is documented as being done by Sendmail. There are two
4936 possible values. For messages generated by the autoreply transport, Exim
4939 Auto-Submitted: auto-replied
4941 whereas for all other generated messages (e.g. bounces) it adds
4943 Auto-Submitted: auto-generated
4945 63. The "personal" condition in filters now includes a test for the
4946 Auto-Submitted: header. If it contains the string "auto-" the message it
4947 not considered personal.
4949 64. Added rcpt_include_affixes as a generic transport option.
4951 65. Added queue_only_override (default true).
4953 66. Added the syslog_duplication option.
4955 67. If what should have been the first header line of a message consisted of
4956 a space followed by a colon, Exim was mis-interpreting it as a header line.
4957 It isn't of course - it is syntactically invalid and should therefore be
4958 treated as the start of the message body. The misbehaviour could have
4959 caused a number of strange effects, including loss of data in subsequent
4960 header lines, and spool format errors.
4962 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4963 client host had authenticated. This control can now be exercised by an ACL
4964 for more flexibility.
4966 69. By default, callouts do not happen when testing with -bh. There is now a
4967 variant, -bhc, which does actually run the callout code, including
4968 consulting and updating the callout cache.
4970 70. Added support for saslauthd authentication, courtesy of Alexander
4973 71. If statvfs() failed on the spool or log directories while checking their
4974 size for availability, Exim confusingly gave the error "space shortage".
4975 Furthermore, in debugging mode it crashed with a floating point exception.
4976 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4977 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4978 serious problem, Exim now writes to the main and panic logs when this
4979 happens, with details of the failure. It then refuses to accept the
4980 incoming message, giving the message "spool directory problem" or "log
4981 directory problem" with a 421 code for SMTP messages.
4983 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4984 0, 1, and 2 exist, because some OS complain for execs without them (see
4985 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4986 descriptors. However, the code omitted to check that the open succeeded,
4987 causing mysterious errors if for some reason the permissions on /dev/null
4988 got screwed. Now Exim writes a message to the main and panic logs, and
4989 bombs out if it can't open /dev/null.
4991 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4992 interact so that it is all more flexible. It is supposed to remain
4993 backwards compatible. Also added extra_local_interfaces.
4995 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4996 to bomb out with an assertion failure - to the client this appears as a
4997 connection drop. This problem occurs in the part of the code that was taken
4998 from the Samba project. Fortunately, the assertion is in a very simple
4999 function, so I have fixed this by reproducing the function inline in the
5000 one place where it is called, and arranging for authentication to fail
5001 instead of killing the process with assert().
5003 75. The SPA client code was not working when the server requested OEM rather
5004 than Unicode encoding.
5006 76. Added code to make require_files with a specific uid setting more usable in
5007 the case where statting the file as root fails - usually a non-root-mounted
5008 NFS file system. When this happens and the failure is EACCES, Exim now
5009 forks a subprocess and does the per-uid checking as the relevant uid.
5011 77. Added process_log_path.
5013 78. If log_file_path was not explicitly set, a setting of check_log_space or
5014 check_log_inodes was ignored.
5016 79. If a space check for the spool or log partitions fails, the incident is now
5017 logged. Of course, in the latter case the data may get lost...
5019 80. Added the %p formatting code to string_format() so that it can be used to
5020 print addresses in debug_print(). Adjusted all the address printing in the
5021 debugging in store.c to use %p rather than %d.
5023 81. There was a concern that a line of code in smtp_in.c could overflow a
5024 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5025 initially expressed, the concern was not well-founded, because trailing
5026 spaces are removed early. However, if the trailing spaces were followed by
5027 a NULL, they did not get removed, so the overflow was possible. Two fixes
5030 (a) I re-wrote the offending code in a cleaner fashion.
5031 (b) If an incoming SMTP command contains a NULL character, it is rejected
5034 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5035 runs initgroups(), so that if the Exim user is in any additional groups,
5036 they will be used during message reception.
5042 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob