1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.606 2010/03/05 16:28:04 nm4 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)
13 NM/01 Bugzilla 965: Support mysql stored procedures.
14 Patch from Alain Williams
16 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
18 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
19 Patch from Andreas Metzler
21 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
22 Patch from Kirill Miazine
24 NM/05 Bugzilla 671: Added umask to procmail example.
30 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body
32 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
34 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
36 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
38 NM/04 Bugzilla 915: Flag broken perl installation during build.
44 TK/01 Added patch by Johannes Berg that expands the main option
45 "spamd_address" if it starts with a dollar sign.
47 TK/02 Write list of recipients to X-Envelope-Sender header when building
48 the mbox-format spool file for content scanning (suggested by Jakob
51 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
52 (http://www.dcc-servers.net/) support via dccifd. Activated by
53 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
55 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
56 by Mark Daniel Reidel <mr@df.eu>.
58 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
59 When building exim an external PCRE library is now needed -
60 PCRE is a system library on the majority of modern systems.
61 See entry on PCRE_LIBS in EDITME file.
63 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
64 conversation. Added nologin parameter to request.
65 Patch contributed by Kirill Miazine
67 TF/01 Do not log submission mode rewrites if they do not change the address.
69 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
71 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
72 log files in place. Contributed by Roberto Lima
74 NM/04 Bugzilla 667: close socket used by dovecot authenticator
76 TF/03 Bugzilla 615: When checking the local_parts router precondition
77 after a local_part_suffix or local_part_prefix option, Exim now
78 does not use the address's named list lookup cache, since this
79 contains cached lookups for the whole local part.
81 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
82 Robert Millan. Documentation is in experimental-spec.txt
84 TF/04 Bugzilla 668: Fix parallel build (make -j).
86 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000
88 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
89 Patch provided by Jan Srzednicki
91 TF/05 Leading white space used to be stripped from $spam_report which
92 wrecked the formatting. Now it is preserved.
94 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
95 that they are available at delivery time.
97 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
99 TF/08 TLS error reporting now respects the incoming_interface and
100 incoming_port log selectors.
102 TF/09 Produce a more useful error message if an SMTP transport's hosts
103 setting expands to an empty string.
105 NM/06 Bugzilla 744: EXPN did not work under TLS.
106 Patch provided by Phil Pennock
108 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
109 Patch provided by Richard Godbee
111 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
112 acl_smtp_notquit, added index entry.
114 NM/09 Bugzilla 787: Potential buffer overflow in string_format
115 Patch provided by Eugene Bujak
117 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
118 Patch provided by Maxim Dounin
120 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
121 Patch provided by Phil Pennock
123 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
125 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
126 Patch provided by Brad "anomie" Jorsch
128 NM/14 Bugzilla 622: Added timeout setting to transport filter.
129 Patch provided by Dean Brooks
131 TK/05 Add native DKIM support (does not depend on external libraries).
133 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
134 Patch provided by Graeme Fowler
136 NM/16 Bugzilla 851: Documentation example syntax fix.
138 NM/17 Changed NOTICE file to remove references to embedded PCRE.
140 NM/18 Bugzilla 894: Fix issue with very long lines including comments in lsearch
142 NM/19 Bugzilla 745: TLS version reporting
143 Patch provided by Phil Pennock
145 NM/20 Bugzilla 167: bool: condition support
146 Patch provided by Phil Pennock
148 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken clients
149 Patch provided by Phil Pennock
151 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date
152 Patch provided by Brad "anomie" Jorsch
154 NM/23 Bugzilla 687: Fix misparses in eximstats
155 Patch provided by Heiko Schlittermann
157 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid
158 Patch provided by Heiko Schlittermann
160 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file
161 plus update to original patch.
163 NM/26 Bugzilla 799: Documentation correction for ratelimit
165 NM/27 Bugzilla 802: Improvements to local interface IP addr detection
166 Patch provided by David Brownlee
168 NM/28 Bugzilla 807: Improvements to LMTP delivery logging
170 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes
172 NM/30 Bugzilla 888: TLS documentation bugfixes
174 NM/31 Bugzilla 896: Dovecot buffer overrun fix
176 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
177 Unlike the original bugzilla I have changed all shell scripts in src tree
179 NM/33 Bugzilla 898: Transport filter timeout fix
180 Patch by Todd Rinaldo
182 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches
183 Patch by Serge Demonchaux
185 NM/35 Bugzilla 39: Base64 decode bug fixes
186 Patch by Jakob Hirsch
188 NM/36 Bugzilla 909: Correct connect() call in dcc code
190 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling
192 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed
194 NM/39 Bugzilla 911: Fixed MakeLinks build script
200 TK/01 Add preliminary DKIM support. Currently requires a forked version of
201 ALT-N's libdkim that I have put here:
202 http://duncanthrax.net/exim-experimental/
204 Note to Michael Haardt: I had to rename some vars in sieve.c. They
205 were called 'true' and it seems that C99 defines that as a reserved
206 keyword to be used with 'bool' variable types. That means you could
207 not include C99-style headers which use bools without triggering
208 build errors in sieve.c.
210 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
211 as mailq or other aliases. Changed the --help handling significantly
212 to do whats expected. exim_usage() emits usage/help information.
214 SC/01 Added the -bylocaldomain option to eximstats.
216 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
218 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
220 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
226 PH/01 Another patch from the Sieve maintainer.
228 PH/02 When an IPv6 address is converted to a string for single-key lookup
229 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
230 dots are used instead of colons so that keys in lsearch files need not
231 contain colons. This was done some time before quoting was made available
232 in lsearch files. However, iplsearch files do require colons in IPv6 keys
233 (notated using the quote facility) so as to distinguish them from IPv4
234 keys. This meant that lookups for IP addresses in host lists did not work
235 for iplsearch lookups.
237 This has been fixed by arranging for IPv6 addresses to be expressed with
238 colons if the lookup type is iplsearch. This is not incompatible, because
239 previously such lookups could never work.
241 The situation is now rather anomolous, since one *can* have colons in
242 ordinary lsearch keys. However, making the change in all cases is
243 incompatible and would probably break a number of configurations.
245 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
248 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
249 conversion specification without a maximum field width, thereby enabling
250 a rogue spamd server to cause a buffer overflow. While nobody in their
251 right mind would setup Exim to query an untrusted spamd server, an
252 attacker that gains access to a server running spamd could potentially
253 exploit this vulnerability to run arbitrary code as the Exim user.
255 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
256 $primary_hostname instead of what libspf2 thinks the hosts name is.
258 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
259 a directory entry by the name of the lookup key. Previously, if a
260 symlink pointed to a non-existing file or a file in a directory that
261 Exim lacked permissions to read, a lookup for a key matching that
262 symlink would fail. Now it is enough that a matching directory entry
263 exists, symlink or not. (Bugzilla 503.)
265 PH/03 The body_linecount and body_zerocount variables are now exported in the
268 PH/04 Added the $dnslist_matched variable.
270 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
271 This means they are set thereafter only if the connection becomes
274 PH/06 Added the client_condition to authenticators so that some can be skipped
275 by clients under certain conditions.
277 PH/07 The error message for a badly-placed control=no_multiline_responses left
278 "_responses" off the end of the name.
280 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
282 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
283 (without spaces) instead of just copying the configuration text.
285 PH/10 Added the /noupdate option to the ratelimit ACL condition.
287 PH/11 Added $max_received_linelength.
289 PH/12 Added +ignore_defer and +include_defer to host lists.
291 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
292 way in which PCRE > 7.0 is built.
294 PH/14 Implemented queue_only_load_latch.
296 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
297 MAIL command. The effect was to mangle the value on 64-bit systems.
299 PH/16 Another patch from the Sieve maintainer.
301 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
303 PH/18 If a system quota error occurred while trying to create the file for
304 a maildir delivery, the message "Mailbox is full" was not appended to the
305 bounce if the delivery eventually timed out. Change 4.67/27 below applied
306 only to a quota excession during the actual writing of the file.
308 PH/19 It seems that peer DN values may contain newlines (and other non-printing
309 characters?) which causes problems in log lines. The DN values are now
310 passed through string_printing() before being added to log lines.
312 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
313 and InterBase are left for another time.)
315 PH/21 Added message_body_newlines option.
317 PH/22 Guard against possible overflow in moan_check_errorcopy().
319 PH/23 POSIX allows open() to be a macro; guard against that.
321 PH/24 If the recipient of an error message contained an @ in the local part
322 (suitably quoted, of course), incorrect values were put in $domain and
323 $local_part during the evaluation of errors_copy.
329 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
330 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
333 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
334 issue a MAIL command.
336 PH/02 In an ACL statement such as
338 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
340 if a client was not listed at all, or was listed with a value other than
341 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
342 the condition was not true (as it should be), so access was not denied.
343 The bug was that the ! inversion was incorrectly passed on to the second
344 item. This has been fixed.
346 PH/03 Added additional dnslists conditions == and =& which are different from
347 = and & when the dns lookup returns more than one IP address.
349 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
350 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
352 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
353 FSYNC, which compiles an option called disable_fsync that allows for
354 bypassing fsync(). The documentation is heavily laced with warnings.
356 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
358 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
359 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
360 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
361 including adding "make clean"; (3) Added -fPIC when compiling the test
362 dynamically loaded module, to get rid of a warning.
364 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
365 message fails, move_frozen_messages = true and ignore_bounce_errors_after
366 = 0s. The bug is otherwise harmless.
368 PH/07 There was a bug in the dovecot authenticator such that the value of
369 $auth1 could be overwritten, and so not correctly preserved, after a
370 successful authentication. This usually meant that the value preserved by
371 the server_setid option was incorrect.
373 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
375 PH/09 Installed PCRE release 7.0.
377 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
378 run for batched SMTP input. It is now run at the start of every message
379 in the batch. While fixing this I discovered that the process information
380 (output by running exiwhat) was not always getting set for -bs and -bS
381 input. This is fixed, and it now also says "batched" for BSMTP.
383 PH/11 Added control=no_pipelining.
385 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
386 patch, slightly modified), and move the expansion of helo_data till after
387 the connection is made in the smtp transport (so it can use these
390 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
392 PH/14 Added log_selector = +pid.
394 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
396 PH/16 Add ${if forany and ${if forall.
398 PH/17 Added dsn_from option to vary the From: line in DSNs.
400 PH/18 Flush SMTP output before performing a callout, unless control =
401 no_callout_flush is set.
403 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
404 was true (the default) a successful delivery failed to delete the retry
405 item, thus causing premature timeout of the address. The bug is now
408 PH/20 Added hosts_avoid_pipelining to the smtp transport.
410 PH/21 Long custom messages for fakedefer and fakereject are now split up
411 into multiline reponses in the same way that messages for "deny" and
412 other ACL rejections are.
414 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
415 with slight modification.
417 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
418 draft, changing the syntax and factoring some duplicate code".
420 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
421 for deliveries of the second and subsequent messages over the same SMTP
424 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
425 ${reduce, with only minor "tidies".
427 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
429 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
430 expansion side effects.
432 PH/27 When a message times out after an over-quota error from an Exim-imposed
433 quota, the bounce message says "mailbox is full". This message was not
434 being given when it was a system quota that was exceeded. It now should
437 MH/03 Made $recipients available in local_scan(). local_scan() already has
438 better access to the recipient list through recipients_list[], but
439 $recipients can be useful in postmaster-provided expansion strings.
441 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
442 in the case of a MAIL command with additional options following the
443 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
444 were accidentally chopped off.
446 PH/29 SMTP synchronization checks are implemented when a command is read -
447 there is a check that no more input is waiting when there shouldn't be
448 any. However, for some commands, a delay in an ACL can mean that it is
449 some time before the response is written. In this time, more input might
450 arrive, invalidly. So now there are extra checks after an ACL has run for
451 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
452 pipelining has not been advertised.
454 PH/30 MH's patch to allow iscntrl() characters to be list separators.
456 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
457 returned in the SMTP response when smtp_return_error_details was false.
460 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
461 instead of the C I/O that was originally supplied, because problems were
464 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
465 Exim which did not show up earlier: it was assuming that a call to
466 SSL_CTX_set_info_callback() might give an error value. In fact, there is
467 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
468 was a macro that became an assignment, so it seemed to work. This has
469 changed to a proper function call with a void return, hence the compile
470 error. Exim's code has been fixed.
472 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
475 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
477 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
479 PH/37 If a message is not accepted after it has had an id assigned (e.g.
480 because it turns out to be too big or there is a timeout) there is no
481 "Completed" line in the log. When some messages of this type were
482 selected by exigrep, they were listed as "not completed". Others were
483 picked up by some special patterns. I have improved the selection
484 criteria to be more general.
486 PH/38 The host_find_failed option in the manualroute router can now be set
487 to "ignore", to completely ignore a host whose IP address cannot be
488 found. If all hosts are ignored, the behaviour is controlled by the new
489 host_all_ignored option.
491 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
492 homing or because of multiple MX records with /mx) generated more than
493 one IP address, and the following item turned out to be the local host,
494 all the secondary addresses of the first item were incorrectly removed
495 from the list, along with the local host and any following hosts (which
496 is what is supposed to happen).
498 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
499 whoever called Exim into the -H file. In the case of the daemon it was
500 behaving confusingly. When first started, it used values for whoever
501 started the daemon, but after a SIGHUP it used the Exim user (because it
502 calls itself on a restart). I have changed the code so that it now always
505 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
506 message are rejected with the same error (e.g. no authentication or bad
507 sender address), and a DATA command is nevertheless sent (as can happen
508 with PIPELINING or a stupid MUA), the error message that was given to the
509 RCPT commands is included in the rejection of the DATA command. This is
510 intended to be helpful for MUAs that show only the final error to their
513 PH/42 Another patch from the Sieve maintainer.
515 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
516 Eximstats - Fixed some broken HTML links and added missing column headers
518 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
519 columns for Rejects, Temp Rejects, Ham, and Spam rows.
521 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
523 PH/43 Yet another patch from the Sieve maintainer.
525 PH/44 I found a way to check for a TCP/IP connection going away before sending
526 the response to the final '.' that terminates a message, but only in the
527 case where the client has not sent further data following the '.'
528 (unfortunately, this is allowed). However, in many cases there won't be
529 any further data because there won't be any more messages to send. A call
530 to select() can be used: if it shows that the input is "ready", there is
531 either input waiting, or the socket has been closed. An attempt to read
532 the next input character can distinguish the two cases. Previously, Exim
533 would have sent an OK response which the client would never have see.
534 This could lead to message repetition. This fix should cure that, at
535 least in a lot of common cases.
537 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
538 advertised in response to EHLO.
544 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
545 fixed by 4.65/MH/01 (is this a record?) are fixed:
547 (i) An empty string was always treated as zero by the numeric comparison
548 operators. This behaviour has been restored.
550 (ii) It is documented that the numeric comparison operators always treat
551 their arguments as decimal numbers. This was broken in that numbers
552 starting with 0 were being interpreted as octal.
554 While fixing these problems I realized that there was another issue that
555 hadn't been noticed. Values of message_size_limit (both the global option
556 and the transport option) were treated as octal if they started with 0.
557 The documentation was vague. These values are now always treated as
558 decimal, and I will make that clear in the documentation.
564 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
565 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
568 MH/01 Don't check that the operands of numeric comparison operators are
569 integers when their expansion is in "skipping" mode (fixes bug
570 introduced by 4.64-PH/07).
572 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
573 child addresses, Exim now panics and dies. Previously, because the count
574 is held in a short int, deliveries were likely to be lost. As such a
575 large number of recipients for a single message is ridiculous
576 (performance will be very, very poor), I have chosen to impose a limit
577 rather than extend the field.
583 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
584 leftover -K file (the existence of which was triggered by #402).
585 While we were at it, introduced process PID as part of the -K
586 filename. This should rule out race conditions when creating
589 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
590 processing considerably. Previous code took too long for large mails,
591 triggering a timeout which in turn triggers #401.
593 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
594 in the DK code in transports.c. sendfile() is not really portable,
595 hence the _LINUX specificness.
597 TF/01 In the add_headers option to the mail command in an Exim filter,
598 there was a bug that Exim would claim a syntax error in any
599 header after the first one which had an odd number of characters
602 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
603 callout verification, Exim cached a "reject" for the entire domain. This
604 is correct for most verifications, but it is not correct for a recipient
605 verification with use_sender or use_postmaster set, because in that case
606 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
607 case of MAIL FROM:<> rejection from other early rejections (e.g.
608 rejection of HELO). When verifying a recipient using a non-null MAIL
609 address, the cache is ignored if it shows MAIL FROM:<> rejection.
610 Whatever the result of the callout, the value of the domain cache is
611 left unchanged (for any other kind of callout, getting as far as trying
612 RCPT means that the domain itself is ok).
614 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
617 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
618 manifest itself as EPIPE rather than ECONNECT. When tidying away a
619 session, the daemon ignores ECONNECT errors and logs others; it now
620 ignores EPIPE as well.
622 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
623 (quoted-printable decoding).
625 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
626 later the small subsequent patch to fix an introduced bug.
628 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
630 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
632 PH/08 An error is now given if message_size_limit is specified negative.
634 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
635 to be given (somewhat) arbitrary names.
637 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
640 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
641 miscellaneous code fixes
643 PH/10 Added the log_reject_target ACL modifier to specify where to log
646 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
647 hostname. This is wrong, because it relates to the incoming message (and
648 probably the interface on which it is arriving) and not to the outgoing
649 callout (which could be using a different interface). This has been
650 changed to use the value of the helo_data option from the smtp transport
651 instead - this is what is used when a message is actually being sent. If
652 there is no remote transport (possible with a router that sets up host
653 addresses), $smtp_active_hostname is used.
655 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
656 tweaks were necessary in order to get it to work (see also 21 below):
657 (a) The code assumed that strncpy() returns a negative number on buffer
658 overflow, which isn't the case. Replaced with Exim's string_format()
660 (b) There were several signed/unsigned issues. I just did the minimum
661 hacking in of casts. There is scope for a larger refactoring.
662 (c) The code used strcasecmp() which is not a standard C function.
663 Replaced with Exim's strcmpic() function.
664 (d) The code set only $1; it now sets $auth1 as well.
665 (e) A simple test gave the error "authentication client didn't specify
666 service in request". It would seem that Dovecot has changed its
667 interface. Fortunately there's a specification; I followed it and
668 changed what the client sends and it appears to be working now.
670 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
673 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
674 address A is aliased to B and C, where B exists and C does not. Without
675 -v the output is "A verified" because verification stops after a
676 successful redirection if more than one address is generated. However,
677 with -v the child addresses are also verified. Exim was outputting "A
678 failed to verify" and then showing the successful verification for C,
679 with its parentage. It now outputs "B failed to verify", showing B's
680 parentage before showing the successful verification of C.
682 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
683 look up a TXT record in a specific list after matching in a combined
686 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
687 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
688 they consult the DNS. I had assumed they would set it the way they
689 wanted; and indeed my experiments on Linux seem to show that in some
690 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
691 To be on the safe side, however, I have now made the interface to
692 host_find_byname() similar to host_find_bydns(), with an argument
693 containing the DNS resolver options. The host_find_byname() function now
694 sets these options at its start, just as host_find_bydns() does. The smtp
695 transport options dns_qualify_single and dns_search_parents are passed to
696 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
697 of host_find_byname() use the default settings of RES_DEFNAMES
698 (qualify_single) but not RES_DNSRCH (search_parents).
700 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
701 spool_read_header() do less string testing, by means of a preliminary
702 switch on the second character of optional "-foo" lines. (This is
703 overdue, caused by the large number of possibilities that now exist.
704 Originally there were few.) While I was there, I also converted the
705 str(n)cmp tests so they don't re-test the leading "-" and the first
706 character, in the hope this might squeeze out yet more improvement.
708 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
709 flag allowing group syntax was set by the header_syntax check but not
710 turned off, possible causing trouble later; (2) The flag was not being
711 set at all for the header_verify test, causing "group"-style headers to
712 be rejected. I have now set it in this case, and also caused header_
713 verify to ignore an empty address taken from a group. While doing this, I
714 came across some other cases where the code for allowing group syntax
715 while scanning a header line wasn't quite right (mostly, not resetting
716 the flag correctly in the right place). These bugs could have caused
717 trouble for malformed header lines. I hope it is now all correct.
719 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
720 with the "reply" argument non-NULL. The code, however (which originally
721 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
722 but it didn't always do it. This confused somebody who was copying the
723 code for some other use. I have removed all the tests.
725 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
726 feature that was used to support insecure browsers during the U.S. crypto
727 embargo. It requires special client support, and Exim is probably the
728 only MTA that supported it -- and would never use it because real RSA is
729 always available. This code has been removed, because it had the bad
730 effect of slowing Exim down by computing (never used) parameters for the
731 RSA_EXPORT functionality.
733 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
734 authenticator to fail if there's a tab character in the incoming data
735 (there should never be unless someone is messing about, as it's supposed
736 to be base64-encoded). Also added, on Timo's advice, the "secured" option
737 if the connection is using TLS or if the remote IP is the same as the
738 local IP, and the "valid-client-cert option" if a client certificate has
741 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
742 authenticators. This can be used for authorization after authentication
743 succeeds. (In the case of plaintext, it servers for both authentication
746 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
747 if any retry times were supplied.
749 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
750 connection, where there is no client IP address to check. In this
751 situation, the verify now always succeeds.
753 PH/25 Applied John Jetmore's -Mset patch.
755 PH/26 Added -bem to be like -Mset, but loading a message from a file.
757 PH/27 In a string expansion for a processed (not raw) header when multiple
758 headers of the same name were present, leading whitespace was being
759 removed from all of them, but trailing whitespace was being removed only
760 from the last one. Now trailing whitespace is removed from each header
761 before concatenation. Completely empty headers in a concatenation (as
764 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
765 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
767 PH/29 [Removed. This was a change that I later backed out, and forgot to
768 correct the ChangeLog entry (that I had efficiently created) before
769 committing the later change.]
771 PH/30 Exim was sometimes attempting to deliver messages that had suffered
772 address errors (4xx response to RCPT) over the same connection as other
773 messages routed to the same hosts. Such deliveries are always "forced",
774 so retry times are not inspected. This resulted in far too many retries
775 for the affected addresses. The effect occurred only when there were more
776 hosts than the hosts_max_try setting in the smtp transport when it had
777 the 4xx errors. Those hosts that it had tried were not added to the list
778 of hosts for which the message was waiting, so if all were tried, there
779 was no problem. Two fixes have been applied:
781 (i) If there are any address or message errors in an SMTP delivery, none
782 of the hosts (tried or untried) are now added to the list of hosts
783 for which the message is waiting, so the message should not be a
784 candidate for sending over the same connection that was used for a
785 successful delivery of some other message. This seems entirely
786 reasonable: after all the message is NOT "waiting for some host".
787 This is so "obvious" that I'm not sure why it wasn't done
788 previously. Hope I haven't missed anything, but it can't do any
789 harm, as the worst effect is to miss an optimization.
791 (ii) If, despite (i), such a delivery is accidentally attempted, the
792 routing retry time is respected, so at least it doesn't keep
793 hammering the server.
795 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
796 in ${readsocket because some servers need this prod.
798 PH/32 Added some extra debug output when updating a wait-xxx database.
800 PH/33 The hint "could be header name not terminated by colon", which has been
801 given for certain expansion errors for a long time, was not being given
802 for the ${if def:h_colon_omitted{... case.
804 PH/34 The spec says: "With one important exception, whenever a domain list is
805 being scanned, $domain contains the subject domain." There was at least
806 one case where this was not true.
808 PH/35 The error "getsockname() failed: connection reset by peer" was being
809 written to the panic log as well as the main log, but it isn't really
810 panic-worthy as it just means the connection died rather early on. I have
811 removed the panic log writing for the ECONNRESET error when getsockname()
814 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
815 runs only) independently of the message's sender address. This meant
816 that, if the 4xx error was in fact related to the sender, a different
817 message to the same recipient with a different sender could confuse
818 things. In particualar, this can happen when sending to a greylisting
819 server, but other circumstances could also provoke similar problems.
820 I have changed the default so that the retry time for these errors is now
821 based a combination of the sender and recipient addresses. This change
822 can be overridden by setting address_retry_include_sender=false in the
825 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
826 remote server are returned as part of bounce messages. This was not
827 happening for LMTP over a pipe (the lmtp transport), but now it is the
828 same for both kinds of LMTP.
830 PH/38 Despite being documented as not happening, Exim was rewriting addresses
831 in header lines that were in fact CNAMEs. This is no longer the case.
833 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
834 and queue runs started by the daemon processed all messages. This has
835 been fixed so that -R and -S can now usefully be given with -q<time>.
837 PH/40 Import PCRE release 6.7 (fixes some bugs).
839 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
841 PH/42 Give an error if -q is specified more than once.
843 PH/43 Renamed the variables $interface_address and $interface_port as
844 $received_ip_address and $received_port, to make it clear that these
845 values apply to message reception, and not to the outgoing interface when
846 a message is delivered. (The old names remain recognized, of course.)
848 PH/44 There was no timeout on the connect() call when using a Unix domain
849 socket in the ${readsocket expansion. There now is.
851 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
852 be meaningful with "accept".
854 SC/01 Eximstats V1.43
855 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
857 SC/02 Eximstats V1.44
858 Use a glob alias rather than an array ref in the generated
859 parser. This improves both readability and performance.
861 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
862 Collect SpamAssassin and rejection statistics.
863 Don't display local sender or destination tables unless
864 there is data to show.
865 Added average volumes into the top table text output.
867 SC/04 Eximstats V1.46
868 Collect data on the number of addresses (recipients)
869 as well as the number of messages.
871 SC/05 Eximstats V1.47
872 Added 'Message too big' to the list of mail rejection
873 reasons (thanks to Marco Gaiarin).
875 SC/06 Eximstats V1.48
876 Mainlog lines which have GMT offsets and are too short to
877 have a flag are now skipped.
879 SC/07 Eximstats V1.49 (Alain Williams)
880 Added the -emptyok flag.
882 SC/08 Eximstats V1.50
883 Fixes for obtaining the IP address from reject messages.
885 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
886 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
887 whitesspace changes from 4.64-PH/27
889 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
892 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
893 are found, allow negative numbers in numeric criteria)
895 JJ/06 exipick.20061117.2, added new $message_body_missing variable
897 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
898 to match changes made in 4.64-PH/43
900 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
902 PH/47 Put in an explicit test for a DNS lookup of an address record where the
903 "domain" is actually an IP address, and force a failure. This locks out
904 those revolvers/nameservers that support "A-for-A" lookups, in
905 contravention of the specifications.
907 PH/48 When a host name was looked up from an IP address, and the subsequent
908 forward lookup of the name timed out, the host name was left in
909 $sender_host_name, contrary to the specification.
911 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
912 restricted to single-key lookups, Exim was not diagnosing an error if
913 * or *@ was used with a query-style lookup.
915 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
917 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
918 long ago, but noone interested enough thought of it. Let's just say that
919 the "1.1" means that there are some new functions that weren't there at
920 some point in the past.
922 PH/51 Error processing for expansion failure of helo_data from an smtp
923 transport during callout processing was broken.
925 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
926 tested/used via the -bh/-bhc/-bs options.
928 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
929 bug, fixed in subsequent PCRE releases).
931 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
932 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
934 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
940 SC/01 Use a glob alias rather than an array ref in eximstats generated
941 parser. This improves both readability and performance.
943 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
944 Don't display local sender or destination tables in eximstats unless
945 there is data to show.
946 Added average volumes into the eximstats top table text output.
948 SC/03 Collect data on the number of addresses (recipients) as well
949 as the number of messages in eximstats.
951 TF/01 Correct an error in the documentation for the redirect router. Exim
952 does (usually) call initgroups() when daemonizing.
954 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
955 with consistent privilege compared to when running as a daemon.
957 TF/03 Note in the spec that $authenticated_id is not set for local
958 submissions from trusted users.
960 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
961 Thanks to Dean Brooks <dean@iglou.com> for the patch.
963 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
964 by adding some example configuration directives to the default
965 configuration file. A little bit of work is required to uncomment the
966 directives and define how usernames and passwords are checked, but
967 there is now a framework to start from.
969 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
970 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
971 without this. I don't know how relevant this is to other LDAP libraries.
973 PH/02 Add the verb name to the "unknown ACL verb" error.
975 PH/03 Magnus Holmgren's patch for filter_prepend_home.
977 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
979 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
980 directory not expanded when it should be if an expanded home directory
981 was set for the address (which is overridden by the transport).
983 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
986 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
987 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
988 because it is too late at that time, and has no effect.
990 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
991 security issue with \' (bugzilla #107). I could not use the
992 PQescapeStringConn() function, because it needs a PGconn value as one of
995 PH/08 When testing addresses using -bt, indicate those final addresses that
996 are duplicates that would not cause an additional delivery. At least one
997 person was confused, thinking that -bt output corresponded to deliveries.
998 (Suppressing duplicates isn't a good idea as you lose the information
999 about possibly different redirections that led to the duplicates.)
1001 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1002 systems where poll() doesn't work, in particular OS X.
1004 PH/10 Added more information to debugging output for retry time not reached.
1006 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1007 operations in malware.c.
1009 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1012 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1013 syslog_duplication set false, log lines that would normally be written
1014 both the the main log and to the reject log were not written to syslog at
1017 PH/14 In the default configuration, change the use of "message" in ACL warn
1018 statements to "add_header".
1020 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1021 not followed by a command (e.g. "seen endif").
1023 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1024 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1027 PH/17 Added extra conditions to the default value of delay_warning_condition
1031 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1032 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1033 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1036 The Auto-Submitted: and various List- headers are standardised, whereas I
1037 don't think Precedence: ever was.
1039 PH/18 Refactored debugging code in route_finduser() to show more information,
1040 in particular, the error code if getpwnam() issues one.
1042 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1043 This is apparently needed in addition to the PH/07 change above to avoid
1044 any possible encoding problems.
1046 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1047 but not after initializing Perl.
1049 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1050 output them only if debugging. By default they are written stderr,
1051 apparently, which is not desirable.
1053 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1056 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1059 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1061 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1062 authenticated or an ident call has been made. Suppress the default
1063 values for $authenticated_id and $authenticated_sender (but permit -oMai
1064 and -oMas) when testing with -bh.
1066 PH/24 Re-jigged the order of the tests in the default configuration so that the
1067 tests for valid domains and recipients precede the DNS black list and CSA
1068 tests, on the grounds that those ones are more expensive.
1070 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1071 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1072 This bug exists in every version of Exim that I still have, right back to
1075 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1076 However, an attempt to turn on case-sensitivity in a regex key by
1077 including (?-i) didn't work because the subject string was already
1078 lowercased, and the effects were non-intuitive. It turns out that a
1079 one-line patch can be used to allow (?-i) to work as expected.
1085 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1086 other effects) broke the use of negated acl sub-conditions.
1088 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1091 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1092 "Deny" causes Exim to reject the incoming connection with a 554 error.
1093 Unfortunately, if there is a major crisis, such as a disk failure,
1094 tcp-wrappers gives "deny", whereas what one would like would be some
1095 kind of temporary error. A kludge has been added to help with this.
1096 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1097 554 error is used if errno is still zero or contains ENOENT (which occurs
1098 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1101 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1103 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1104 errors. Otherwise a message that provokes a temporary error (when other
1105 messages do not) can cause a whole host to time out.
1107 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1108 addresses were routed directly to files or pipes from a redirect router.
1109 File deliveries just didn't batch; pipe deliveries might have suffered
1112 PH/06 A failure to get a lock for a hints database would erroneously always say
1113 "Failed to get write lock", even when it was really a read lock.
1115 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1116 of 0600. This has been changed to use the value of the lockfile_mode
1117 option (which defaults to 0600).
1119 PH/08 Applied small patch from the Sieve maintainer.
1121 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1122 folder from quota calculations, a direct delivery into this folder messed
1123 up the contents of the maildirsize file. This was because the regex was
1124 used only to exclude .Trash (or whatever) when the size of the mailbox
1125 was calculated. There was no check that a delivery was happening into an
1126 excluded directory. This bug has been fixed by ignoring all quota
1127 processing for deliveries into excluded directories.
1129 PH/10 Added the maildirfolder_create_regex option to appendfile.
1135 PH/01 The code for finding all the local interface addresses on a FreeBSD
1136 system running IPv6 was broken. This may well have applied to all BSD
1137 systems, as well as to others that have similar system calls. The broken
1138 code found IPv4 interfaces correctly, but gave incorrect values for the
1139 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1140 that it would not match correctly against @[] and not recognize the IPv6
1143 PH/02 The ipliteral router was not recognizing addresses of the form user@
1144 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1146 PH/03 Added disable_ipv6.
1148 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1149 decoded form, because it is most often used to construct To: headers
1150 lines in autoreplies, and the decoded form may well be syntactically
1151 invalid. However, $reply_address has leading white space removed, and all
1152 newlines turned into spaces so that the autoreply transport does not
1155 PH/05 If group was specified without a user on a router, and no group or user
1156 was specified on a transport, the group from the router was ignored.
1158 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1159 for visible compile-time settings that can be used to change these
1160 numbers, for those that want even more. Backwards compatibility with old
1161 spool files has been maintained. However, going back to a previous Exim
1162 release will lost any variables that are in spool files.
1164 PH/07 Two small changes when running in the test harness: increase delay when
1165 passing a TCP/IP connection to a new process, in case the original
1166 process has to generate a bounce, and remove special handling of
1167 127.0.0.2 (sic), which is no longer necessary.
1169 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1170 be the same on different OS.
1172 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1175 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1176 whether --show-vars was specified or not
1178 JJ/02 exipick: Added support for new ACL variable spool format introduced
1181 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1182 syntactically invalid From: or Reply-to: line, and a filter used this to
1183 generate an autoreply, and therefore failed to obtain an address for the
1184 autoreply, Exim could try to deliver to a non-existent relative file
1185 name, causing unrelated and misleading errors. What now happens is that
1186 it logs this as a hard delivery error, but does not attempt to create a
1189 PH/11 The exinext utility has a -C option for testing purposes, but although
1190 the given file was scanned by exinext itself; it wasn't being passed on
1191 when Exim was called.
1193 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1194 an end-of-file indication when reading a command response.
1196 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1197 compiled. In many other places in Exim, IPv6 addresses are always
1198 recognized, so I have changed this. It also means that IPv4 domain
1199 literals of the form [IPV4:n.n.n.n] are now always recognized.
1201 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1202 used if the router is not running as root, for example, when verifying at
1203 ACL time, or when using -bh. The debugging output from this situation was
1204 non-existent - all you got was a failure to exec. I have made two
1207 (a) Failures to set uid/gid, the current directory, or a process leader
1208 in a subprocess such as that created by queryprogram now generate
1209 suitable debugging ouput when -d is set.
1211 (b) The queryprogram router detects when it is not running as root,
1212 outputs suitable debugging information if -d is set, and then runs
1213 the subprocess without attempting to change uid/gid.
1215 PH/15 Minor change to Makefile for building test_host (undocumented testing
1218 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1219 additional section of a DNS packet that returns MX or SRV records.
1220 Instead, it always explicitly searches for A/AAAA records. This avoids
1221 major problems that occur when a DNS server includes only records of one
1222 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1223 fixed another bug: if SRV records were looked up and the corresponding
1224 address records were *not* found in the additional section, the port
1225 values from the SRV records were lost.
1227 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1228 using the correct key (the original address) when searching the retry
1229 rules in order to find which one to use for generating the retry hint.
1231 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1232 adding the default one. Similarly, if it contains a Reply-To: header, the
1233 errors_reply_to option, if set, is not used.
1235 PH/19 When calculating a retry time, Exim used to measure the "time since
1236 failure" by looking at the "first failed" field in the retry record. Now
1237 it does not use this if it is later than than the arrival time of the
1238 message. Instead it uses the arrival time. This makes for better
1239 behaviour in cases where some deliveries succeed, thus re-setting the
1240 "first failed" field. An example is a quota failure for a huge message
1241 when small messages continue to be delivered. Without this change, the
1242 "time since failure" will always be short, possible causing more frequent
1243 delivery attempts for the huge message than are intended.
1244 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1246 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1247 $1, $2, $3) because the numerical variables can be reset during some
1248 expansion items (e.g. "match"), thereby losing the authentication data.
1250 PH/21 Make -bV show the size of off_t variables so that the test suite can
1251 decide whether to run tests for quotas > 2G.
1253 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1254 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1255 filecount value is greater than 2G or if a quota value is greater than 2G
1256 on a system where the size of off_t is not greater than 4, a panic error
1259 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1260 never match. The debug and -bh output now contains an explicit error
1261 message indicating a malformed IPv4 address or mask.
1263 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1264 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1265 PH/23 above applies.
1267 PH/25 Do not write to syslog when running in the test harness. The only
1268 occasion when this arises is a failure to open the main or panic logs
1269 (for which there is an explicit test).
1271 PH/26 Added the /no_tell option to "control=freeze".
1273 PH/27 If a host name lookup failed very early in a connection, for example, if
1274 the IP address matched host_lookup and the reverse lookup yielded a name
1275 that did not have a forward lookup, an error message of the form "no IP
1276 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1277 could be logged. Now it outputs the IP address instead of "NULL".
1279 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1280 allows the sender and the authenticated sender to be set when
1281 submitting a message from within Exim. Since child_open_exim() is
1282 documented for local_scan(), the new function should be too.
1284 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1285 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1286 results in an empty string is now treated as unset.
1288 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1290 PH/31 Added sender_verify_fail logging option.
1292 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1293 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1294 tidied the source and removed it altogether.
1296 PH/33 When a queue run was abandoned because the load average was too high, a
1297 log line was always written; now it is written only if the queue_run log
1298 selector is set. In addition, the log line for abandonment now contains
1299 information about the queue run such as the pid. This is always present
1300 in "start" and "stop" lines but was omitted from the "abandon" line.
1302 PH/34 Omit spaces between a header name and the colon in the error message that
1303 is given when verify = headers_syntax fails (if there are lots of them,
1304 the message gets confusing).
1306 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1307 names, as there are now some PTR records that contain slashes. This check
1308 is only to protect against broken name servers that fall over on strange
1309 characters, so the fact that it applies to all lookups doesn't matter.
1311 PH/36 Now that the new test suite is complete, we can remove some of the
1312 special code in Exim that was needed for the old test suite. For example,
1313 sorting DNS records because real resolvers return them in an arbitrary
1314 order. The new test suite's fake resolver always returns records in the
1317 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1318 bounces) except when queue_only is set, to avoid logging races between
1319 the different processes.
1321 PH/38 Panic-die if .include specifies a non-absolute path.
1323 PH/39 A tweak to the "H" retry rule from its user.
1325 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1326 a label. They prevented compilation on older perls.
1328 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1329 a warning to be raised on newish perls.
1331 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1332 on queue. Changes to match documented behaviour of showing count of
1333 messages matching specified criteria.
1335 PH/40 Changed the default ident timeout from 30s to 5s.
1337 PH/41 Added support for the use of login_cap features, on those BSD systems
1338 that have them, for controlling the resources used by pipe deliveries.
1340 PH/42 The content-scanning code uses fopen() to create files in which to put
1341 message data. Previously it was not paying any attention to the mode of
1342 the files. Exim runs with umask(0) because the rest of the code creates
1343 files with open(), and sets the required mode explicitly. Thus, these
1344 files were ending up world-writeable. This was not a big issue, because,
1345 being within the spool directory, they were not world-accessible. I have
1346 created a function called modefopen, which takes an additional mode
1347 argument. It sets umask(777), creates the file, chmods it to the required
1348 mode, then resets the umask. All the relevant calls to fopen() in the
1349 content scanning code have been changed to use this function.
1351 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1352 to 24 hours. This avoids potential overflow problems when processing G
1353 and H retry rules. I suspect nobody ever tinkers with this value.
1355 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1357 PH/45 When the plaintext authenticator is running as a client, the server's
1358 challenges are checked to ensure they are valid base64 strings. By
1359 default, the authentication attempt is cancelled if an invalid string is
1360 received. Setting client_ignore_invalid_base64 true ignores these errors.
1361 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1362 they are received. Thus, the responses can be made to depend on the
1363 challenges. If an invalid string is ignored, an empty string is placed in
1366 PH/46 Messages that are created by the autoreply transport now contains a
1367 References: header, in accordance with RFCs 2822 and 3834.
1369 PH/47 Added authenticated_sender_force to the smtp transport.
1371 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1373 PH/49 Installed latest patch from the Sieve maintainer.
1375 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1376 was also set, the appendfile transport was unnecessarily scanning a
1377 directory of message files (e.g. for maildir delivery) to find the count
1378 of files (along with the size), even though it did not need this
1379 information. It now does the scan only if it needs to find either the
1380 size of the count of files.
1382 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1384 PH/52 Two bugs concerned with error handling when the smtp transport is
1387 (i) Exim was not creating retry information for temporary errors given
1388 for individual recipients after the DATA command when the smtp transport
1389 was used in LMTP mode. This meant that they could be retried too
1390 frequently, and not timed out correctly.
1392 (ii) Exim was setting the flag that allows error details to be returned
1393 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1394 recipients that were returned after the DATA command.
1396 PH/53 This is related to PH/52, but is more general: for any failing address,
1397 when detailed error information was permitted to be returned to the
1398 sender, but the error was temporary, then after the final timeout, only
1399 "retry timeout exceeded" was returned. Now it returns the full error as
1400 well as "retry timeout exceeded".
1402 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1403 do this, and (what is worse) MTAs that accept it.
1405 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1406 will now be deprecated.
1408 PH/56 New os.c-cygwin from the Cygwin maintainer.
1410 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1411 formats (previously only available in exim formats via -bpr, -bpru,
1412 and -bpra. Now also available in native and exiqgrep formats)
1414 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1415 with very large, slow to parse queues
1417 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1419 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1421 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1422 responses to authentication challenges, though it was showing the
1423 challenges; (ii) I've removed the CR characters from the debug output for
1426 PH/58 Allow for the insertion of a newline as well as a space when a string
1427 is turned into more than one encoded-word during RFC 2047 encoding. The
1428 Sieve code now uses this.
1430 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1431 data_4xx, lost_connection, tls_required.
1433 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1434 message was being sent as an SMTP response.
1436 PH/61 Add -l and -k options to exicyclog.
1438 PH/62 When verifying, if an address was redirected to one new address, so that
1439 verification continued, and the new address failed or deferred after
1440 having set something in $address_data, the value of $address_data was not
1441 passed back to the ACL. This was different to the case when no
1442 redirection occurred. The value is now passed back in both cases.
1444 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1445 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1446 use login_cap.h, so on its own it isn't the distinguishing feature. The
1447 new name refers directly to the setclassresources() function.
1449 PH/65 Added configuration files for NetBSD3.
1451 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1453 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1454 is preferred over IPv4.
1456 PH/68 The bounce_return_message and bounce_return_body options were not being
1457 honoured for bounces generated during the reception of non-SMTP messages.
1458 In particular, this applied to messages rejected by the ACL. This bug has
1459 been fixed. However, if bounce_return_message is true and bounce_return_
1460 body is false, the headers that are returned for a non-SMTP message
1461 include only those that have been read before the error was detected.
1462 (In the case of an ACL rejection, they have all been read.)
1464 PH/69 The HTML version of the specification is now built in a directory called
1465 spec_html instead of spec.html, because the latter looks like a path with
1466 a MIME-type, and this confuses some software.
1468 PH/70 Catch two compiler warnings in sieve.c.
1470 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1471 function verify_get_ident() calls ip_connect() to connect a socket, but
1472 if the "connect()" function timed out, ip_connect() used to close the
1473 socket. However, verify_get_ident() also closes the socket later, and in
1474 between Exim writes to the log, which may get opened at this point. When
1475 the socket was closed in ip_connect(), the log could get the same file
1476 descriptor number as the socket. This naturally causes chaos. The fix is
1477 not to close the socket in ip_connect(); the socket should be closed by
1478 the function that creates it. There was only one place in the code where
1479 this was missing, in the iplookup router, which I don't think anybody now
1480 uses, but I've fixed it anyway.
1482 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1483 well as to direct DNS lookups. Otherwise the handling of names in host
1484 lists is inconsistent and therefore confusing.
1490 PH/01 Two changes to the default runtime configuration:
1492 (1) Move the checks for relay_from_hosts and authenticated clients from
1493 after to before the (commented out) DNS black list checks.
1495 (2) Add control=submission to the relay_from_hosts and authenticated
1496 clients checks, on the grounds that messages accepted by these
1497 statements are most likely to be submissions.
1499 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1501 (1) Generate an error if the third argument for the ${prvs expansion is
1504 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1507 (3) Reset the variables that are obtained from the first argument of
1508 ${prvscheck and used in the second argument before leaving the code,
1509 because their memory is reclaimed, so using them afterwards may do
1512 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1513 one (it's much easier than Tom thought :-).
1515 (5) Because of (4), we can now allow for the use of $prvscheck_result
1516 inside the third argument.
1518 PH/03 For some reason, the default setting of PATH when running a command from
1519 a pipe transport was just "/usr/bin". I have changed it to
1522 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1523 anything to be listed in the output from -bV.
1525 PH/05 When a filter generated an autoreply, the entire To: header line was
1526 quoted in the delivery log line, like this:
1528 => >A.N.Other <ano@some.domain> <original@ddress> ...
1530 This has been changed so that it extracts the operative address. There
1531 may be more than one such address. If so, they are comma-separated, like
1534 => >ano@some.domain,ona@other.domain <original@ddress> ...
1536 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1537 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1538 not being looked up in the rDNS to get a host name, Exim was showing the
1539 IP address twice in Received: lines, even though the IP addresses were
1540 identical. For example:
1542 Received: from [1.2.3.4] (helo=[1.2.3.4])
1544 However, if the real host name was known, it was omitting the HELO data
1545 if it matched the actual IP address. This has been tidied up so that it
1546 doesn't show the same IP address twice.
1548 PH/07 When both +timestamp and +memory debugging was on, the value given by
1549 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1550 called by the debug printing, thereby overwriting the timestamp buffer.
1551 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1553 PH/08 When the original message was included in an autoreply transport, it
1554 always said "this is a copy of the message, including all the headers",
1555 even if body_only or headers_only was set. It now gives an appropriate
1558 PH/09 Applied a patch from the Sieve maintainer which:
1560 o fixes some comments
1561 o adds the (disabled) notify extension core
1562 o adds some debug output for the result of if/elsif tests
1563 o points to the current vacation draft in the documentation
1564 and documents the missing references header update
1568 o fixes a bug in processing the envelope test (when testing
1569 multiple envelope elements, the last element determinted the
1572 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1573 Electronic Mail") by including:
1575 Auto-submitted: auto-generated
1577 in the messages that it generates (bounce messages and others, such as
1578 warnings). In the case of bounce messages for non-SMTP mesages, there was
1579 also a typo: it was using "Auto_submitted" (underscore instead of
1580 hyphen). Since every message generated by Exim is necessarily in response
1581 to another message, thes have all been changed to:
1583 Auto-Submitted: auto-replied
1585 in accordance with these statements in the RFC:
1587 The auto-replied keyword:
1589 - SHOULD be used on messages sent in direct response to another
1590 message by an automatic process,
1592 - MUST NOT be used on manually-generated messages,
1594 - MAY be used on Delivery Status Notifications (DSNs) and Message
1595 Disposition Notifications (MDNs),
1597 - MUST NOT be used on messages generated by automatic or periodic
1598 processes, except for messages which are automatic responses to
1601 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1602 to the default Received: header definition.
1604 PH/12 Added log selector acl_warn_skipped (default on).
1606 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1607 variables because (a) they are in the wrong storage pool and (b) even if
1608 they were copied, it wouldn't work properly because of the caching.
1610 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1611 checking when decoding. Apparently there are clients that generate
1612 overlong encoded strings. Why am I not surprised?
1614 PH/15 If the first argument of "${if match_address" was not empty, but did not
1615 contain an "@" character, Exim crashed. Now it writes a panic log message
1616 and treats the condition as false.
1618 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1620 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1621 extension "envelope-auth". The code is finished and in agreement with
1622 other implementations, but there is no documentation so far and in fact,
1623 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1624 not changing the active code.
1626 Print executed "if" and "elsif" statements when debugging is used. This
1627 helps a great deal to understand what a filter does.
1629 Document more things not specified clearly in RFC3028. I had all this
1630 sorted out, when out of a sudden new issues came to my mind. Oops."
1632 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1635 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1636 canonical form (as documented). However, after a host name lookup from
1637 the IP address, check_host() was doing a simple string comparison with
1638 addresses acquired from the DNS when checking that the found name did
1639 have the original IP as one of its addresses. Since any found IPv6
1640 addresses are likely to be in abbreviated form, the comparison could
1641 fail. Luckily, there already exists a function for doing the comparison
1642 by converting both addresses to binary, so now that is used instead of
1643 the text comparison.
1645 PH/20 There was another similar case to PH/19, when a complete host name was
1646 given in a host list; looking up its IP address could give an abbreviated
1647 form, whereas the current host's name might or might not be abbreviated.
1648 The same fix has been applied.
1654 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1655 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1658 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1659 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1661 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1663 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1664 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1665 Also, for "Auto-Submitted", treat anything other than "no" as
1666 non-personal, in accordance with RFC 3834. (Previously it treated
1667 anything starting "auto-" as non-personal.)
1669 TF/01 The control=submission/name=... option had a problem with syntax
1670 errors if the name included a slash character. The /name= option
1671 now slurps the rest of the string, so it can include any characters
1672 but it must come last in the list of options (after /sender_retain
1675 PH/05 Some modifications to the interface to the fake nameserver for the new
1683 TK/01 Added the "success_on_redirect" address verification option. See
1684 NewStuff for rationale and an example.
1686 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1688 PH/02 Patch to exigrep to allow it to work on syslog lines.
1690 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1691 fread() to skip over the body file's header line, because in Cygwin the
1692 header line is locked and is inaccessible.
1694 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1695 co-exist for some time) to make it clear that it is the Exim ID that is
1696 referenced, not the Message-ID: header line.
1698 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1699 string_format() function, because snprintf() does not exist on all
1702 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1703 ${stat: expansion item.
1705 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1706 protocol synchronization error", to keep the pedants happy.
1708 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1709 well as for IRIX systems, when gcc is being used. See the host.c source
1712 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1714 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1717 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1718 transport and to the smtp transport in LMTP mode.
1720 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1722 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1723 run a filter in a subprocess. This could lead to confusion in subsequent
1724 lookups in the parent process. There should also be a search_tidyup() at
1725 the end of the subprocess.
1727 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1728 only if the host matched helo_try_verify_hosts, which caused the
1729 verification to occur when the EHLO/HELO command was issued. The ACL just
1730 tested the remembered result. Now, if a previous verification attempt has
1731 not happened, "verify = helo" does it there and then.
1733 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1735 TK/03 Fix log output including CR from clamd.
1737 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1738 exist provoked a memory error which could cause a segfault.
1740 PH/15 Installed PCRE 6.2
1742 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1744 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1745 of the problem. Specifically, suggested +O2 rather than +O1 for the
1748 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1750 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1751 record, the port was not correctly logged when the outgoing_port log
1752 selector was set (it logged the transort's default port).
1754 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1755 fallback_hosts, and "hosts" in the smtp transport.
1757 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1758 host errors such as "Connection refused".
1760 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1761 authentication with radiusclient 0.4.9:
1763 - Error returned from rc_read_config was caught wrongly
1764 - Username/password not passed on to radius server due to wrong length.
1766 The presumption is that some radiusclient API changes for 4.51/PH/17
1767 were not taken care of correctly. The code is still untested by me (my
1768 Linux distribution still has 0.3.2 of radiusclient), but it was
1769 contributed by a Radius user.
1771 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1772 expanding the "port" option of the smtp transport.
1774 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1775 while reading a MIME header. Thanks to Tom Hughes for a patch.
1777 PH/24 Include config.h inside local_scan.h so that configuration settings are
1780 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1781 that in an ACL for RCPT (for example), you can examine exactly what was
1784 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1785 commands, but it was not correctly comparing the address with the actual
1786 client host address. Thus, it would show the EHLO address in Received:
1787 header lines when this was not necessary.
1789 PH/27 Added the % operator to ${eval:}.
1791 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1792 it should be ignoring failures (because with -C, for example, it has lost
1793 privilege). It wasn't ignoring creation failures other than "already
1796 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1799 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1800 because an input line was too long, either on its own, or by virtue of
1801 too many continuations, the temporary file was not being removed, and the
1802 return code was incorrect.
1804 PH/31 Missing "BOOL" in function definition in filtertest.c.
1806 PH/32 Applied Sieve patches from the maintainer.
1808 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1810 PH/33 Added "verify = not_blind".
1812 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1813 Local/Makefile (with some defaults set). These are used in built scripts
1814 such as exicyclog, but they have never been used in the exim_install
1815 script (though there are many overriding facilities there). I have
1816 arranged that the exim_install script now takes note of these two
1819 PH/35 Installed configuration files for Dragonfly.
1821 PH/36 When a locally submitted message by a trusted user did not contain a
1822 From: header, and the sender address was obtained from -f or from an SMTP
1823 MAIL command, and the trusted user did not use -F to supply a sender
1824 name, $originator_name was incorrectly used when constructing a From:
1825 header. Furthermore, $originator_name was used for submission mode
1826 messages from external hosts without From: headers in a similar way,
1827 which is clearly wrong.
1829 PH/37 Added control=suppress_local_fixups.
1831 PH/38 When log_selector = +received_sender was set, and the addition of the
1832 sender made the log line's construction buffer exactly full, or one byte
1833 less than full, an overflow happened when the terminating "\n" was
1836 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1837 when the result of a list match is failure because a DNS lookup failed.
1839 PH/40 RM_COMMAND is now used in the building process.
1841 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1842 the "build-* directories that it finds.
1844 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1845 address in a domain literal was a prefix of an interface address.
1847 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1848 when verifying a sender address, unless rewrite_headers is false.
1850 PH/44 Wrote a long comment about why errors_to addresses are verified as
1851 recipients, not senders.
1853 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1854 the ratelimit ACL was added.
1856 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1858 PH/46 Added extra information about PostgreSQL errors to the error string.
1860 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1861 avoiding the need to install special zones in a real server. This is
1862 backwards compatible; if it can't find the fake resolver, it drops back.
1863 Thus, both old and new test suites can be run.
1865 TF/02 Added util/ratelimit.pl
1867 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1868 clock is set back in time.
1870 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1871 Candler <B.Candler@pobox.com>.
1873 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1874 OK for addresses that are the result of redirections.
1876 PH/48 A number of further additions for the benefit of the new test suite,
1877 including a fake gethostbyname() that interfaces to the fake DNS resolver
1880 TF/06 The fix for widen_domains has also been applied to qualify_single and
1881 search_parents which are the other dnslookup options that can cause
1884 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1887 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1889 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1890 service" if it encountered an MX record with an empty target hostname.
1891 The message is now "an MX or SRV record indicated no SMTP service".
1893 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1894 if the DNS of the sending site is misconfigured. This is quite a
1895 common situation. This change restores the behaviour of treating a
1896 helo verification defer as a failure.
1898 PH/51 If self=fail was set on a router, the bounce message did not include the
1899 actual error message.
1905 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1907 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1908 command itself ended in error, the underlying message about the transport
1909 filter timeout was being overwritten with the pipe command error. Now the
1910 underlying error message should be appended to the second error message.
1912 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1914 PH/02 Reduce the amount of output that "make" produces by default. Full output
1915 can still be requested.
1917 PH/03 The warning log line about a condition test deferring for a "warn" verb
1918 was being output only once per connection, rather than after each
1919 occurrence (because it was using the same function as for successful
1920 "warn" verbs). This seems wrong, so I have changed it.
1922 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1923 it should not have, which might have caused a crash in the right
1924 circumstances, but probably never did.
1926 PH/04 Installed a modified version of Tony Finch's patch to make submission
1927 mode fix the return path as well as the Sender: header line, and to
1928 add a /name= option so that you can make the user's friendly name appear
1931 TF/03 Added the control = fakedefer ACL modifier.
1933 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1934 Mark Lowes for thorough testing.
1936 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1938 TK/03 Merged latest SRS patch from Miles Wilton.
1940 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1941 with the definition in sysexits.h (which is #included earlier).
1942 Fortunately, Exim does not actually use EX_OK. The code used to try to
1943 preserve the sysexits.h value, by assumimg that macro definitions were
1944 scanned for macro replacements. I have been disabused of this notion,
1945 so now the code just undefines EX_OK before #including unistd.h.
1947 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1948 in the smtp transport. When a block could not be written in a single
1949 write() function, the timeout was being re-applied to each part-write.
1950 This seems wrong - if the receiver was accepting one byte at a time it
1951 would take for ever. The timeout is now adjusted when this happens. It
1952 doesn't have to be particularly precise.
1954 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1955 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1957 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1958 without a domain if the check to <postmaster@domain> fails.
1960 SC/01 Eximstats: added -xls and the ability to specify output files
1961 (patch written by Frank Heydlauf).
1963 SC/02 Eximstats: use FileHandles for outputing results.
1965 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1967 SC/04 Eximstats: fixed display of large numbers with -nvr option
1969 SC/05 Eximstats: fixed merging of reports with empty tables.
1971 SC/06 Eximstats: added the -include_original_destination flag
1973 SC/07 Eximstats: removed tabs and trailing whitespace.
1975 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1977 TK/06 MBOX spool code: Add real "From " MBOX separator line
1978 so the .eml file is really in mbox format (even though
1979 most programs do not really care). Patch from Alex Miller.
1981 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1982 The latter is generated from $received_to and is only set if the
1983 message has one envelope recipient. SA can use these headers,
1984 obviously out-of-the-box. Patch from Alex Miller.
1986 PH/08 The ${def test on a variable was returning false if the variable's
1987 value was "0", contrary to what the specification has always said!
1988 The result should be true unless the variable is empty.
1990 PH/09 The syntax error of a character other than { following "${if
1991 def:variable_name" (after optional whitespace) was not being diagnosed.
1992 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1993 accidental colon was present, for example, could give incorrect results.
1995 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1996 result is used (not including appendfile, where other changes are about
1999 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2000 This involved changing a lot of size variables from int to off_t. It
2001 should work with maildirs and everything.
2003 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2004 spamd dying while we are connected to it.
2006 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2009 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2010 function for BDB 4.3.
2012 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2014 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2015 the sources for PCRE have been re-arranged and more files are now
2018 PH/15 The code I had for printing potentially long long variables in PH/11
2019 above was not the best (it lost precision). The length of off_t variables
2020 is now inspected at build time, and an appropriate printing format (%ld
2021 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2022 to be "long long int" or "long int". This is needed for the internal
2023 formatting function string_vformat().
2025 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2026 the configuration file to be ":syslog", then the script "guesses" where
2027 the logs files are, rather than using the compiled in default. In our
2028 case the guess is not the same as the compiled default, so the script
2029 suddenly stopped working when I started to use syslog. The patch checks
2030 to see if log_file_path is "". If so, it attempts to read it from exim
2031 with no configuration file to get the compiled in version, before it
2032 falls back to the previous guessing code."
2034 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2035 implementing BATV in an Exim configuration. See NewStuff for the gory
2038 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2039 Makefile that are specific to HP-UX.
2041 PH/18 If the "use_postmaster" option was set for a recipient callout together
2042 with the "random" option, the postmaster address was used as the MAIL
2043 FROM address for the random test, but not for the subsequent recipient
2044 test. It is now used for both.
2046 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2047 patch removes a few documentation additions to RFC 3028, because the
2048 latest draft now contains them. It adds the new en;ascii-case comparator
2049 and a new error check for 8bit text in MIME parts. Comparator and
2050 require names are now matched exactly. I enabled the subaddress
2051 extension, but it is not well tested yet (read: it works for me)."
2053 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2054 rework some of the code of TK/09 above to avoid the hardwired use of
2055 "%lld" and "long long". Replaced the call to snprintf() with a call to
2058 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2059 records point to non-existent hosts", "retry timeout exceeded", and
2060 "retry time not reached for any host after a long failure period".
2062 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2063 experimental DomainKeys support:
2065 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2066 (2) On an error such as an illegally used "control", the wrong name for
2067 the control was given.
2069 These problems did NOT occur unless DomainKeys support was compiled.
2071 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2073 PH/24 Added ${if match_ip condition.
2075 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2076 databases so that it will be absolutely obvious if a crash occurs in the
2077 DB library. This is a regular occurrence (often caused by mis-matched
2080 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2081 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2082 picked up on a user's system that detects such things. There doesn't seem
2083 to be a gcc warning option for this - only an attribute that has to be
2084 put on the function's prototype. It seems that in Fedora Core 4 they have
2085 set this on a number of new functions. No doubt there will be more in due
2088 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2089 specify a transport. However, if an address that was verified by such a
2090 router was the subject of a callout, Exim crashed because it tried to
2091 read the rcpt_include_affixes from the non-existent transport. Now it
2092 just assumes that the setting of that option is false. This bug was
2093 introduced by 4.51/PH/31.
2095 PH/28 Changed -d+all to exclude +memory, because that information is very
2096 rarely of interest, but it makes the output a lot bigger. People tend to
2097 do -d+all out of habit.
2099 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2100 code in os-type was giving problems when libc.so lives in lib64, like on
2103 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2104 aren't the modern standard, and it seems that some systems' include files
2105 don't always have them. Exim was already checking for some of the newer
2106 ones like T_AAAA, and defining it itself. I've added checks for all the
2107 record types that Exim uses.
2109 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2110 not automatically generating a new one, as it is supposed to. This
2111 prevented TLS from working. If the file did exist, but contained invalid
2112 data, a new version was generated, as expected. It was only the case of a
2113 non-existent file that was broken.
2115 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2116 with a change in libdomainkeys > 0.64.
2118 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2119 from DNS. If the selector record carries the flag, it now has
2120 precedence over the domain-wide flag.
2122 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2124 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2125 the use of a port name that isn't defined in /etc/services) occurred, the
2126 message was deferred as in a normal delivery, and thus remained on the
2127 spool, instead of being failed because of the mua_wrapper setting. This
2128 is now fixed, and I tidied up some of the mua_wrapper messages at the
2131 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2132 the messages in a hash of arrays rather than using individual hashes.
2133 This is a bit cleaner and results in dramatic memory savings, albeit
2134 at a slight CPU cost.
2136 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2137 as requested by Marc Sherman.
2139 SC/10 Eximstats: added histograms for user specified patterns as requested
2142 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2144 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2145 fopen() in the content-scanning modules that did not already have it.
2151 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2152 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2154 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2156 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2158 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2159 to test Sieve filters that use "vacation".
2161 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2162 that changes the way the GnuTLS parameters are stored in the cache file.
2163 The new format can be generated externally. For backward compatibility,
2164 if the data in the cache doesn't make sense, Exim assumes it has read an
2165 old-format file, and it generates new data and writes a new file. This
2166 means that you can't go back to an older release without removing the
2169 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2170 work if there are any delivery delays because "one_time" forces the
2171 parent to be marked "delivered", so its unseen clone is never tried
2172 again. For this reason, Exim now forbids the simultaneous setting of
2175 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2176 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2177 entry if you want to know the details. The fix, however, neglected to
2178 consider the case when local delivery batching is involved. The test for
2179 "previously delivered" was not happening when checking to see if an
2180 address could be batched with a previous (undelivered) one; under
2181 certain circumstances this could lead to multiple deliveries to the same
2184 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2185 in its include files, and this causes problems building Exim.
2187 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2188 header_syntax) but Exim was just ignoring anything given after a slash.
2189 In particular, this caused confusion with an attempt to use "verify =
2190 reverse_host_lookup/defer_ok". An error is now given when options are
2191 supplied for verify items that do not have them. (Maybe reverse_host_
2192 lookup should have a defer_ok option, but that's a different point.)
2194 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2195 defined by RFC 821) to 2048, because there were problems with some AUTH
2196 commands, and RFC 1869 says the size should be increased for extended
2197 SMTP commands that take arguments.
2199 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2202 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2203 "unknown" error; now it says that the functionality isn't in the binary.
2205 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2206 an address' error message when a string expansion fails (syntax or
2207 whatever). Otherwise the password may appear in the log. Following change
2208 PH/42 below, there is no longer a chance of it appearing in a bounce
2211 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2213 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2214 of its addresses was ever tried. (Bugzilla bug #2.)
2216 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2217 the result incorrectly in the debug output. (It correctly added a newline
2218 to what was transported.)
2220 TF/01 Added $received_time.
2222 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2223 commented out examples of how to interface to a virus scanner and to
2224 SpamAssassin. Also added commented examples of av_scanner and
2225 spamd_address settings.
2227 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2228 and controls are allowed in which ACLs. There were a couple of minor
2229 errors. Some of the entries in the conditions table (which is a table of
2230 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2231 negation of where the condition IS allowed.
2233 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2235 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2236 header file does not have a version number, so I've had to invent a new
2237 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2238 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2239 radiusclient), but it was contributed by a Radius user.
2241 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2242 files or directories, for OpenSSL.
2244 PH/19 When an Exim process that is running as root has to create an Exim log
2245 file, it does so in a subprocess that runs as exim:exim so as to get the
2246 ownership right at creation (otherwise, other Exim processes might see
2247 the file with the wrong ownership). There was no test for failure of this
2248 fork() call, which would lead to the process getting stuck as it waited
2249 for a non-existent subprocess. Forks do occasionally fail when resources
2250 run out. I reviewed all the other calls to fork(); they all seem to check
2253 PH/20 When checking for unexpected SMTP input at connect time (before writing
2254 the banner), Exim was not dealing correctly with a non-positive return
2255 from the read() function. If the client had disconnected by this time,
2256 the result was a log entry for a synchronization error with an empty
2257 string after "input=" when read() returned zero. If read() returned -1
2258 (an event I could not check), uninitialized data bytes were printed.
2259 There were reports of junk text (parts of files, etc) appearing after
2262 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2264 PH/22 Added support for macro redefinition, and (re)definition in between
2265 driver and ACL definitions.
2267 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2268 forgetting to use the resulting value; it was using the unexpanded value.
2270 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2271 hadn't been configured. The fix is from Juergen Kreileder, who
2272 understands it better than I do:
2274 "Here's what I see happening with three configured cyrus_sasl
2275 authenticators configured (plain, login, cram-md5):
2277 On startup auth_cyrus_sasl_init() gets called for each of these.
2278 This means three calls to sasl_listmech() without a specified mech_list.
2279 => SASL tests which mechs of all available mechs actually work
2280 => three warnings about OTP not working
2281 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2283 With the patch, sasl_listmech() also gets called three times. But now
2284 SASL's mech_list option is set to the server_mech specified in the the
2285 authenticator. Or in other words, the answer from sasl_listmech()
2286 gets limited to just the mech you're testing for (which is different
2288 => the return list contains just 'plain' or 'login', 'cram-md5' or
2289 nothing depending on the value of ob->server_mech.
2291 I've just tested the patch: Authentication still works fine,
2292 unavailable mechs specified in the exim configuration are still
2293 caught, and the auth.log warnings about OTP are gone."
2295 PH/25 When debugging is enabled, the contents of the command line are added
2296 to the debugging output, even when log_selector=+arguments is not
2299 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2300 answer is "GNU", and only if the return is "GNU/something" is the answer
2303 PH/27 $acl_verify_message is now set immediately after the failure of a
2304 verification in an ACL, and so is available in subsequent modifiers. In
2305 particular, the message can be preserved by coding like this:
2307 warn !verify = sender
2308 set acl_m0 = $acl_verify_message
2310 Previously, $acl_verify_message was set only while expanding "message"
2311 and "log_message" when a very denied access.
2313 PH/28 Modified OS/os.c-Linux with
2315 -#ifndef OS_LOAD_AVERAGE
2316 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2318 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2319 nomenclature these days.)
2321 PH/29 Installed patch from the Sieve maintainer that adds the options
2322 sieve_useraddress and sieve_subaddress to the redirect router.
2324 PH/30 In these circumstances:
2325 . Two addresses routed to the same list of hosts;
2326 . First host does not offer TLS;
2327 . First host accepts first address;
2328 . First host gives temporary error to second address;
2329 . Second host offers TLS and a TLS session is established;
2330 . Second host accepts second address.
2331 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2332 and peerdn, if requested) that were in fact used only for the second
2335 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2336 attention to any local part prefix or suffix that was matched by the
2337 router that accepted the address. It now behaves in the same way as it
2338 does for delivery: the affixes are removed from the local part unless
2339 rcpt_include_affixes is set on the transport.
2341 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2342 timeout during the DATA phase of an incoming message.
2344 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2345 applied a patch sanctioned by the Sieve maintainer.
2347 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2348 the uid or gid is negative. A case of a negative gid caused this to be
2349 noticed. The fix allows for either to be negative.
2351 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2352 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2355 PH/36 The change PH/12 above was broken. Fixed it.
2357 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2358 the grounds that routing the same address twice would always produce the
2359 same answer. This might have been true once, but it is certainly no
2360 longer true now. Routing a child address may depend on the previous
2361 routing that produced that child. Some complicated redirection strategies
2362 went wrong when messages had multiple recipients, and made Exim's
2363 behaviour dependent on the order in which the addresses were given.
2365 I have moved the duplicate checking until after the routing is complete.
2366 Exim scans the addresses that are assigned to local and remote
2367 transports, and removes any duplicates. This means that more work will be
2368 done, as duplicates will always all be routed, but duplicates are
2369 presumably rare, so I don't expect this is of any significance.
2371 For deliveries to pipes, files, and autoreplies, the duplicate checking
2372 still happens during the routing process, since they are not going to be
2375 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2376 It corrects a timeout issue with spamd. This is Ian's comment: "The
2377 background is that sometimes spamd either never reads data from a
2378 connection it has accepted, or it never writes response data. The exiscan
2379 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2380 blindly assumes that writes won't block so it may never time out."
2382 PH/39 Allow G after quota size as well as K and M.
2384 PH/40 The value set for $authenticated_id in an authenticator may not contain
2385 binary zeroes or newlines because the value is written to log lines and
2386 to spool files. There was no check on this. Now the value is run through
2387 the string_printing() function so that such characters are converted to
2388 printable escape sequences.
2390 PH/41 $message_linecount is a new variable that contains the total number of
2391 lines in the message. Compare $body_linecount, which is the count for the
2394 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2395 bounce and delay warning messages, except in certain special cases, which
2398 (a) An SMTP error message from a remote host;
2399 (b) A message specified in a :fail: redirection;
2400 (c) A message specified in a "fail" command in a system filter;
2401 (d) A message specified in a FAIL return from the queryprogram router;
2402 (e) A message specified by the cannot_route_message router option.
2404 In these cases only, Exim does include the error details in bounce and
2405 warning messages. There are also a few cases where bland messages such
2406 as "unrouteable address" or "local delivery error" are given.
2408 PH/43 $value is now also set for the "else" part of a ${run expansion.
2410 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2411 being worked on, but at least Exim now implements the latest version to
2414 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2415 process to complete was treated as a delivery failure, a timeout while
2416 writing the message to the pipe was logged, but erroneously treated as a
2417 successful delivery. Such timeouts include transport filter timeouts. For
2418 consistency with the overall process timeout, these timeouts are now
2419 treated as errors, giving rise to delivery failures by default. However,
2420 there is now a new Boolean option for the pipe transport called
2421 timeout_defer, which, if set TRUE, converts the failures into defers for
2422 both kinds of timeout. A transport filter timeout is now identified in
2425 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2426 systems where "make" and "gmake" are different, calling "gmake" at top
2427 level broke things. I've arranged for the value of $(MAKE) to be passed
2428 from the Makefile to this script so that it can call the same version of
2432 A note about Exim versions 4.44 and 4.50
2433 ----------------------------------------
2435 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2436 changes of various kinds. As a consequence, a big documentation update was
2437 needed. This delayed the release for rather longer than seemed good, especially
2438 in the light of a couple of (minor) security issues. Therefore, the changes
2439 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2440 release. So 4.44 and 4.50 are in effect two different branches that both start
2443 I have left the 4.50 change log unchanged; it contains all the changes since
2444 4.43. The change log for 4.44 is below; many of its items are identical to
2445 those for 4.50. This seems to be the most sensible way to preserve the
2446 historical information.
2452 1. Minor wording change to the doc/README.SIEVE file.
2454 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2455 computation of the current number of files was incorrect.
2457 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2458 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2461 4. Give more explanation in the error message when the command for a transport
2462 filter fails to execute.
2464 5. There are several places where Exim runs a non-Exim command in a
2465 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2466 was being done only for the command run by the queryprogram router. It is
2467 now done for all such subprocesses. The other cases are: ${run, transport
2468 filters, and the commands run by the lmtp and pipe transports.
2470 6. Added CONFIGURE_GROUP build-time option.
2472 7. Some older OS have a limit of 256 on the maximum number of file
2473 descriptors. Exim was using setrlimit() to set 1000 as a large value
2474 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2475 systems. I've change it so that if it can't get 1000, it tries for 256.
2477 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2478 was an oversight, and furthermore, ever since the addition of extra
2479 controls (e.g. 4.43/32), the checks on when to allow different forms of
2480 "control" were broken. There should now be diagnostics for all cases when a
2481 control that does not make sense is encountered.
2483 9. Added the /retain_sender option to "control=submission".
2485 10. $recipients is now available in the predata ACL (oversight).
2487 11. Tidy the search cache before the fork to do a delivery from a message
2488 received from the command line. Otherwise the child will trigger a lookup
2489 failure and thereby defer the delivery if it tries to use (for example) a
2490 cached ldap connection that the parent has called unbind on.
2492 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2493 of $address_data from the recipient verification was clobbered by the
2494 sender verification.
2496 13. The value of address_data from a sender verification is now available in
2497 $sender_address_data in subsequent conditions in the ACL statement.
2499 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2501 15. Added a new option "connect=<time>" to callout options, to set a different
2504 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2505 was its contents. (It was OK if the option was not defined at all.)
2507 17. A "Completed" log line is now written for messages that are removed from
2508 the spool by the -Mrm option.
2510 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2511 information about exactly what failed.
2513 19. Added -dd to debug only the daemon process.
2515 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2516 handles timeouts, both on the server side and network timeouts. Renamed the
2517 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2519 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2520 It is now set to "smtps".
2522 22. $host_address is now set to the target address during the checking of
2523 ignore_target_hosts.
2525 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2526 being passed; this would have caused $sender_host_name to have been used if
2527 matching the list had actually called for a host name (not very likely,
2528 since this list is usually IP addresses). A host name is now passed as
2531 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2532 code that specifies a non-restarting handler (typically sigaction() in
2533 modern systems) in an attempt to fix a rare and obscure crash bug.
2535 25. Narrowed the window for a race in the daemon that could cause it to ignore
2536 SIGCHLD signals. This is not a major problem, because they are used only to
2537 wake it up if nothing else does.
2539 26. A malformed maildirsize file could cause Exim to calculate negative values
2540 for the mailbox size or file count. Odd effects could occur as a result.
2541 The maildirsize information is now recalculated if the size or filecount
2544 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2545 support for a long time. Removed HAVE_SYS_VFS_H.
2547 28. Installed the latest version of exipick from John Jetmore.
2549 29. In an address list, if the pattern was not a regular expression, an empty
2550 subject address (from a bounce message) matched only if the pattern was an
2551 empty string. Non-empty patterns were not even tested. This was the wrong
2552 because it is perfectly reasonable to use an empty address as part of a
2553 database query. An empty address is now tested by patterns that are
2554 lookups. However, all the other forms of pattern expect the subject to
2555 contain a local part and a domain, and therefore, for them, an empty
2556 address still always fails if the pattern is not itself empty.
2558 30. Exim went into a mad DNS loop when attempting to do a callout where the
2559 host was specified on an smtp transport, and looking it up yielded more
2560 than one IP address.
2562 31. Re-factored the code for checking spool and log partition space into a
2563 function that finds that data and another that does the check. The former
2564 is then used to implement four new variables: $spool_space, $log_space,
2565 $spool_inodes, and $log_inodes.
2567 32. The RFC2047 encoding function was originally intended for short strings
2568 such as real names; it was not keeping to the 75-character limit for
2569 encoded words that the RFC imposes. It now respects the limit, and
2570 generates multiple encoded words if necessary. To be on the safe side, I
2571 have increased the buffer size for the ${rfc2047: expansion operator from
2574 33. It is now permitted to omit both strings after an "if" condition; if the
2575 condition is true, the result is "true". As before, when the second string
2576 is omitted, a false condition yields an empty string. This makes it less
2577 cumbersome to write custom ACL and router conditions.
2579 34. Failure to deliver a bounce message always caused it to be frozen, even if
2580 there was an errors_to setting on the router. The errors_to setting is now
2583 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2584 canonical form (fully expanded) before being placed in
2585 $sender_host_address.
2587 36. The table in the code that translates DNS record types into text (T_A to
2588 "A" for instance) was missing entries for NS and CNAME. It is just possible
2589 that this could have caused confusion if both these types were looked up
2590 for the same domain, because the text type is used as part of Exim's
2591 per-process caching. But the chance of anyone hitting this buglet seems
2594 37. The dnsdb lookup has been extended in a number of ways.
2596 (1) There is a new type, "zns", which walks up the domain tree until it
2597 finds some nameserver records. It should be used with care.
2599 (2) There is a new type, "mxh", which is like "mx" except that it returns
2600 just the host names, not the priorities.
2602 (3) It is now possible to give a list of domains (or IP addresses) to be
2603 looked up. The behaviour when one of the lookups defers can be
2604 controlled by a keyword.
2606 (4) It is now possible to specify the separator character for use when
2607 multiple records are returned.
2609 38. The dnslists ACL condition has been extended: it is now possible to supply
2610 a list of IP addresses and/or domains to be looked up in a particular DNS
2613 39. Added log_selector=+queue_time_overall.
2615 40. When running the queue in the test harness, wait just a tad after forking a
2616 delivery process, to get repeatability of debugging output.
2618 41. Include certificate and key file names in error message when GnuTLS fails
2619 to set them up, because the GnuTLS error message doesn't include the name
2620 of the failing file when there is a problem reading it.
2622 42. Allow both -bf and -bF in the same test run.
2624 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2626 44. The "Exiscan patch" is now merged into the mainline Exim source.
2628 45. Sometimes the final signoff response after QUIT could fail to get
2629 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2630 < 0 before doing a fflush(). This bug looks as though it goes back to the
2631 introduction of TLS in release 3.20, but "sometimes" must have been rare
2632 because the tests only now provoked it.
2634 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2635 (this can affect the format of dates).
2637 47. exim_tidydb, when checking for the continued existence of a message for
2638 which it has found a message-specific retry record, was not finding
2639 messages that were in split spool directories. Consequently, it was
2640 deleting retry records that should have stayed in existence.
2642 48. Steve fixed some bugs in eximstats.
2644 49. The SPA authentication driver was not abandoning authentication and moving
2645 on to the next authenticator when an expansion was forced to fail,
2646 contradicting the general specification for all authenticators. Instead it
2647 was generating a temporary error. It now behaves as specified.
2649 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2650 (the order specifies the preference for clients). The order is now AES256,
2651 AES128, 3DES, ARCFOUR128.
2653 51. Small patch to Sieve code - explicitly set From: when generating an
2656 52. Exim crashed if a remote delivery caused a very long error message to be
2657 recorded - for instance if somebody sent an entire SpamAssassin report back
2658 as a large number of 550 error lines. This bug was coincidentally fixed by
2659 increasing the size of one of Exim's internal buffers (big_buffer) that
2660 happened as part of the Exiscan merge. However, to be on the safe side, I
2661 have made the code more robust (and fixed the comments that describe what
2664 53. Now that there can be additional text after "Completed" in log lines (if
2665 the queue_time_overall log selector is set), a one-byte patch to exigrep
2666 was needed to allow it to recognize "Completed" as not the last thing in
2669 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2670 patch that reportedly fixes this has been added. I am not expert enough to
2671 create a test for it. This is what the patch creator wrote:
2673 "I found a little strange behaviour of ldap code when working with
2674 Windows 2003 AD Domain, where users was placed in more than one
2675 Organization Units. When I tried to give exim partial DN, the exit code
2676 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2677 But simultaneously result of request was absolutely normal ldap result,
2678 so I produce this patch..."
2680 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2681 REFERENCE, so I have modified the code to exclude the patch when that macro
2684 55. Some experimental protocols are using DNS PTR records for new purposes. The
2685 keys for these records are domain names, not reversed IP addresses. The
2686 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2687 leaves it alone. Component reversal etc. now happens only for IP addresses.
2690 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2692 57. Double the size of the debug message buffer (to 2048) so that more of very
2693 long debug lines gets shown.
2695 58. The exicyclog utility now does better if the number of log files to keep
2696 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2698 59. Two changes related to the smtp_active_hostname option:
2700 (1) $smtp_active_hostname is now available as a variable.
2701 (2) The default for smtp_banner uses $smtp_active_hostname instead
2702 of $primary_hostname.
2704 60. The host_aton() function is supposed to be passed a string that is known
2705 to be a valid IP address. However, in the case of IPv6 addresses, it was
2706 not checking this. This is a hostage to fortune. Exim now panics and dies
2707 if the condition is not met. A case was found where this could be provoked
2708 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2709 components; fortuitously, this particular loophole had already been fixed
2710 by change 4.50/55 above.
2712 If there are any other similar loopholes, the new check in host_aton()
2713 itself should stop them being exploited. The report I received stated that
2714 data on the command line could provoke the exploit when Exim was running as
2715 exim, but did not say which command line option was involved. All I could
2716 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2717 running as the user.
2720 61. There was a buffer overflow vulnerability in the SPA authentication code
2721 (which came originally from the Samba project). I have added a test to the
2722 spa_base64_to_bits() function which I hope fixes it.
2725 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2726 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2728 63. The daemon start-up calls getloadavg() while still root for those OS that
2729 need the first call to be done as root, but it missed one case: when
2730 deliver_queue_load_max is set with deliver_drop_privilege. This is
2731 necessary for the benefit of the queue runner, because there is no re-exec
2732 when deliver_drop_privilege is set.
2734 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2735 This has been fixed.
2737 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2738 was in use, was not putting the data itself into the right store pool;
2739 consequently, it could be overwritten for a subsequent message in the same
2740 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2743 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2745 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2746 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2747 the function were treating the return as a boolean value, which happened to
2748 work because 0=false and not-0=true, but is not correct code.
2750 68. The host_aton() function was not handling scoped IPv6 addresses (those
2751 with, for example, "%eth0" on the end) correctly.
2753 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2754 negated items (that is, ~something) in unsigned ints. Some compilers
2755 apparently mutter when there is no cast.
2757 70. If an address verification called from an ACL failed, and did not produce a
2758 user-specific message (i.e. there was only a "system" message), nothing was
2759 put in $acl_verify_message. In this situation, it now puts the system
2762 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2763 session; change 4.31/43 added the unwanted input to the log line - except
2764 that it did not do this in the start of session case. It now does.
2766 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2767 This is wrong and can cause the other end to generate a synchronization
2768 error if it is another Exim or anything else that does the synchronization
2769 check. A QUIT command is no longer sent after a timeout.
2771 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2772 during host lookups.
2774 74. The defer_ok option of callout verification was not working if it was used
2775 when verifying addresses in header lines, that is, for this case:
2777 verify = header_sender/callout=defer_ok
2779 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2780 those file descriptors could be used for SMTP connections. If anything
2781 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2782 could be sent to the SMTP client, causing chaos. The daemon now opens
2783 stdin, stdout, and stderr to /dev/null when it puts itself into the
2786 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2787 log by default. The user can override this with suitable Perl magic.
2789 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2790 the log message when discard triggers, was not working for the DATA ACL or
2791 for the non-SMTP ACL.
2793 78. Error message wording change in sieve.c.
2795 79. If smtp_accept_max_per_host was set, the number of connections could be
2796 restricted to fewer than expected, because the daemon was trying to set up
2797 a new connection before checking whether the processes handling previous
2798 connections had finished. The check for completed processes is now done
2799 earlier. On busy systems, this bug wouldn't be noticed because something
2800 else would have woken the daemon, and it would have reaped the completed
2803 80. If a message was submitted locally by a user whose login name contained one
2804 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2805 It caused a spool format error. I have fixed the spool reading code. A
2806 related problem was that the "from" clause in the Received: line became
2807 illegal because of the space(s). It is now covered by ${quote_local_part.
2809 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2812 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2814 83. Patch from Sieve maintainer to fix unterminated string problem in
2815 "vacation" handling.
2817 84. Some minor changes to the Linux configuration files to help with other
2818 OS variants using glibc.
2820 85. One more patch for Sieve to update vacation handling to latest spec.
2823 ----------------------------------------------------
2824 See the note above about the 4.44 and 4.50 releases.
2825 ----------------------------------------------------
2831 1. Change 4.43/35 introduced a bug that caused file counts to be
2832 incorrectly computed when quota_filecount was set in an appendfile
2835 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2836 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2839 3. Give more explanation in the error message when the command for a transport
2840 filter fails to execute.
2842 4. There are several places where Exim runs a non-Exim command in a
2843 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2844 was being done only for the command run by the queryprogram router. It is
2845 now done for all such subprocesses. The other cases are: ${run, transport
2846 filters, and the commands run by the lmtp and pipe transports.
2848 5. Some older OS have a limit of 256 on the maximum number of file
2849 descriptors. Exim was using setrlimit() to set 1000 as a large value
2850 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2851 systems. I've change it so that if it can't get 1000, it tries for 256.
2853 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2854 was an oversight, and furthermore, ever since the addition of extra
2855 controls (e.g. 4.43/32), the checks on when to allow different forms of
2856 "control" were broken. There should now be diagnostics for all cases when a
2857 control that does not make sense is encountered.
2859 7. $recipients is now available in the predata ACL (oversight).
2861 8. Tidy the search cache before the fork to do a delivery from a message
2862 received from the command line. Otherwise the child will trigger a lookup
2863 failure and thereby defer the delivery if it tries to use (for example) a
2864 cached ldap connection that the parent has called unbind on.
2866 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2867 of $address_data from the recipient verification was clobbered by the
2868 sender verification.
2870 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2871 was its contents. (It was OK if the option was not defined at all.)
2873 11. A "Completed" log line is now written for messages that are removed from
2874 the spool by the -Mrm option.
2876 12. $host_address is now set to the target address during the checking of
2877 ignore_target_hosts.
2879 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2880 being passed; this would have caused $sender_host_name to have been used if
2881 matching the list had actually called for a host name (not very likely,
2882 since this list is usually IP addresses). A host name is now passed as
2885 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2886 code that specifies a non-restarting handler (typically sigaction() in
2887 modern systems) in an attempt to fix a rare and obscure crash bug.
2889 15. Narrowed the window for a race in the daemon that could cause it to ignore
2890 SIGCHLD signals. This is not a major problem, because they are used only to
2891 wake it up if nothing else does.
2893 16. A malformed maildirsize file could cause Exim to calculate negative values
2894 for the mailbox size or file count. Odd effects could occur as a result.
2895 The maildirsize information is now recalculated if the size or filecount
2898 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2899 support for a long time. Removed HAVE_SYS_VFS_H.
2901 18. Updated exipick to current release from John Jetmore.
2903 19. Allow an empty sender to be matched against a lookup in an address list.
2904 Previously the only cases considered were a regular expression, or an
2907 20. Exim went into a mad DNS lookup loop when doing a callout where the
2908 host was specified on the transport, if the DNS lookup yielded more than
2911 21. The RFC2047 encoding function was originally intended for short strings
2912 such as real names; it was not keeping to the 75-character limit for
2913 encoded words that the RFC imposes. It now respects the limit, and
2914 generates multiple encoded words if necessary. To be on the safe side, I
2915 have increased the buffer size for the ${rfc2047: expansion operator from
2918 22. Failure to deliver a bounce message always caused it to be frozen, even if
2919 there was an errors_to setting on the router. The errors_to setting is now
2922 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2923 canonical form (fully expanded) before being placed in
2924 $sender_host_address.
2926 24. Updated eximstats to version 1.33
2928 25. Include certificate and key file names in error message when GnuTLS fails
2929 to set them up, because the GnuTLS error message doesn't include the name
2930 of the failing file when there is a problem reading it.
2932 26. Expand error message when OpenSSL has problems setting up cert/key files.
2935 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2936 (this can affect the format of dates).
2938 28. exim_tidydb, when checking for the continued existence of a message for
2939 which it has found a message-specific retry record, was not finding
2940 messages that were in split spool directories. Consequently, it was
2941 deleting retry records that should have stayed in existence.
2943 29. eximstats updated to version 1.35
2944 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2945 1.35 - bugfix such that pie charts by volume are generated correctly
2947 30. The SPA authentication driver was not abandoning authentication and moving
2948 on to the next authenticator when an expansion was forced to fail,
2949 contradicting the general specification for all authenticators. Instead it
2950 was generating a temporary error. It now behaves as specified.
2952 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2953 (the order specifies the preference for clients). The order is now AES256,
2954 AES128, 3DES, ARCFOUR128.
2956 31. Small patch to Sieve code - explicitly set From: when generating an
2959 32. Exim crashed if a remote delivery caused a very long error message to be
2960 recorded - for instance if somebody sent an entire SpamAssassin report back
2961 as a large number of 550 error lines. This bug was coincidentally fixed by
2962 increasing the size of one of Exim's internal buffers (big_buffer) that
2963 happened as part of the Exiscan merge. However, to be on the safe side, I
2964 have made the code more robust (and fixed the comments that describe what
2967 33. Some experimental protocols are using DNS PTR records for new purposes. The
2968 keys for these records are domain names, not reversed IP addresses. The
2969 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2970 leaves it alone. Component reversal etc. now happens only for IP addresses.
2973 34. The host_aton() function is supposed to be passed a string that is known
2974 to be a valid IP address. However, in the case of IPv6 addresses, it was
2975 not checking this. This is a hostage to fortune. Exim now panics and dies
2976 if the condition is not met. A case was found where this could be provoked
2977 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2978 components; fortuitously, this particular loophole had already been fixed
2979 by change 4.50/55 or 4.44/33 above.
2981 If there are any other similar loopholes, the new check in host_aton()
2982 itself should stop them being exploited. The report I received stated that
2983 data on the command line could provoke the exploit when Exim was running as
2984 exim, but did not say which command line option was involved. All I could
2985 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2986 running as the user.
2989 35. There was a buffer overflow vulnerability in the SPA authentication code
2990 (which came originally from the Samba project). I have added a test to the
2991 spa_base64_to_bits() function which I hope fixes it.
2994 36. The daemon start-up calls getloadavg() while still root for those OS that
2995 need the first call to be done as root, but it missed one case: when
2996 deliver_queue_load_max is set with deliver_drop_privilege. This is
2997 necessary for the benefit of the queue runner, because there is no re-exec
2998 when deliver_drop_privilege is set.
3000 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3001 was in use, was not putting the data itself into the right store pool;
3002 consequently, it could be overwritten for a subsequent message in the same
3003 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3006 38. Sometimes the final signoff response after QUIT could fail to get
3007 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3008 < 0 before doing a fflush(). This bug looks as though it goes back to the
3009 introduction of TLS in release 3.20, but "sometimes" must have been rare
3010 because the tests only now provoked it.
3016 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3017 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3018 Now it returns an integer. A number of calls were still expecting a T/F
3019 return. Fortuitously, in all cases, the tests worked in OK situations,
3020 which is the norm. However, things would have gone wrong on any write
3021 failures on the smtp file descriptor. This function is used when sending
3022 messages over SMTP and also when doing verify callouts.
3024 2. When Exim is called to do synchronous delivery of a locally submitted
3025 message (the -odf or -odi options), it no longer closes stderr before doing
3028 3. Implemented the mua_wrapper option.
3030 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3032 5. Implemented the functions header_remove(), header_testname(),
3033 header_add_at_position(), and receive_remove_recipient(), and exported them
3036 6. If an ACL "warn" statement specified the addition of headers, Exim already
3037 inserted X-ACL-Warn: at the start if there was no header name. However, it
3038 was not making this test for the second and subsequent header lines if
3039 there were newlines in the string. This meant that an invalid header could
3040 be inserted if Exim was badly configured.
3042 7. Allow an ACL "warn" statement to add header lines at the start or after all
3043 the Received: headers, as well as at the end.
3045 8. Added the rcpt_4xx retry error code.
3047 9. Added postmaster_mailfrom=xxx to callout verification option.
3049 10. Added mailfrom=xxxx to the callout verification option, for verify=
3052 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3053 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3055 12. Inserted some casts to stop certain compilers warning when using pointer
3056 differences as field lengths or precisions in printf-type calls (mostly
3057 affecting debugging statements).
3059 13. Added optional readline() support for -be (dynamically loaded).
3061 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3062 same clock tick as a message's arrival, so that its received time was the
3063 same as the "first fail" time on the retry record, and that message
3064 remained on the queue past the ultimate address timeout, every queue runner
3065 would try a delivery (because it was past the ultimate address timeout) but
3066 after another failure, the ultimate address timeout, which should have then
3067 bounced the address, did not kick in. This was a "< instead of <=" error;
3068 in most cases the first failure would have been in the next clock tick
3069 after the received time, and all would be well.
3071 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3072 being recognized when the domain list was tested by the match_domain
3073 condition in an expansion string.
3075 16. Added the ${str2b64: operator.
3077 17. Exim was always calling setrlimit() to set a large limit for the number of
3078 processes, without checking whether the existing limit was already
3079 adequate. (It did check for the limit on file descriptors.) Furthermore,
3080 errors from getrlimit() and setrlimit() were being ignored. Now they are
3081 logged to the main and panic logs, but Exim does carry on, to try to do its
3082 job under whatever limits there are.
3084 18. Imported PCRE 5.0.
3086 19. Trivial typo in log message " temporarily refused connection" (the leading
3089 20. If the log selector return_path_on_delivery was set and an address was
3090 redirected to /dev/null, the delivery process crashed because it assumed
3091 that a return path would always be set for a "successful" delivery. In this
3092 case, the whole delivery is bypassed as an optimization, and therefore no
3095 21. Internal re-arrangement: the function for sending a challenge and reading
3096 a response while authentication was assuming a zero-terminated challenge
3097 string. It's now changed to take a pointer and a length, to allow for
3098 binary data in such strings.
3100 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3102 23. Exim was not respecting finduser_retries when seeking the login of the
3103 uid under which it was called; it was always trying 10 times. (The default
3104 setting of finduser_retries is zero.) Also, it was sleeping after the final
3105 failure, which is pointless.
3107 24. Implemented tls_on_connect_ports.
3109 25. Implemented acl_smtp_predata.
3111 26. If the domain in control=submission is set empty, Exim assumes that the
3112 authenticated id is a complete email address when it generates From: or
3113 Sender: header lines.
3115 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3116 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3117 chown and chgrp in /bin and hostname in /usr/bin.
3119 28. Exim was keeping the "process log" file open after each use, just as it
3120 does for the main log. This opens the possibility of it remaining open for
3121 long periods when the USR1 signal hits a daemon. Occasional processlog
3122 errors were reported, that could have been caused by this. Anyway, it seems
3123 much more sensible not to leave this file open at all, so that is what now
3126 29. The long-running daemon process does not normally write to the log once it
3127 has entered its main loop, and it closes the log before doing so. This is
3128 so that log files can straightforwardly be renamed and moved. However,
3129 there are a couple of unusual error situations where the daemon does write
3130 log entries, and I had neglected to close the log afterwards.
3132 30. The text of an SMTP error response that was received during a remote
3133 delivery was being truncated at 512 bytes. This is too short for some of
3134 the long messages that one sometimes sees. I've increased the limit to
3137 31. It is now possible to make retry rules that apply only when a message has a
3138 specific sender, in particular, an empty sender.
3140 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3141 it possible to be selective about when SMTP synchronization is enforced.
3143 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3145 32. Implemented hosts_connection_nolog.
3147 33. Added an ACL for QUIT.
3149 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3152 35. Added mailbox_size and mailbox_filecount to appendfile.
3154 36. Added control = no_multiline_responses to ACLs.
3156 37. There was a bug in the logic of the code that waits for the clock to tick
3157 in the case where the clock went backwards by a substantial amount such
3158 that the microsecond fraction of "now" was more than the microsecond
3159 fraction of "then" (but the whole seconds number was less).
3161 38. Added support for the libradius Radius client library this is found on
3162 FreeBSD (previously only the radiusclient library was supported).
3168 1. When certain lookups returned multiple values in the form name=value, the
3169 quoting of the values was not always being done properly. Specifically:
3170 (a) If the value started with a double quote, but contained no whitespace,
3172 (b) If the value contained whitespace other than a space character (i.e.
3173 tabs or newlines or carriage returns) it was not quoted.
3174 This fix has been applied to the mysql and pgsql lookups by writing a
3175 separate quoting function and calling it from the lookup code. The fix
3176 should probably also be applied to nisplus, ibase and oracle lookups, but
3177 since I cannot test any of those, I have not disturbed their existing code.
3179 2. A hit in the callout cache for a specific address caused a log line with no
3180 reason for rejecting RCPT. Now it says "Previous (cached) callout
3181 verification failure".
3183 3. There was an off-by-one bug in the queryprogram router. An over-long
3184 return line was truncated at 256 instead of 255 characters, thereby
3185 overflowing its buffer with the terminating zero. As well as fixing this, I
3186 have increased the buffer size to 1024 (and made a note to document this).
3188 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3189 when Exim is waiting for an SMTP response from a remote server, Exim
3190 restarts its select() call on the socket, thereby resetting its timeout.
3191 This is not a problem when such interrupts are rare. Somebody set up a cron
3192 job to run exiwhat every 2 minutes, which is less than the normal select()
3193 timeout (5 or 10 minutes). This meant that the select() timeout never
3194 kicked in because it was always reset. I have fixed this by comparing the
3195 time when an interrupt arrives with the time at the start of the first call
3196 to select(). If more time than the timeout has elapsed, the interrupt is
3197 treated as a timeout.
3199 5. Some internal re-factoring in preparation for the addition of Sieve
3200 extensions (by MH). In particular, the "personal" test is moved to a
3201 separate function, and given an option for scanning Cc: and Bcc: (which is
3202 not set for Exim filters).
3204 6. When Exim created an email address using the login of the caller as the
3205 local part (e.g. when creating a From: or Sender: header line), it was not
3206 quoting the local part when it contained special characters such as @.
3208 7. Installed new OpenBSD configuration files.
3210 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3211 try to make them clearer.
3213 9. Callout options, other than the timeout value, were being ignored when
3214 verifying sender addresses in header lines. For example, when using
3216 verify = header_sender/callout=no_cache
3218 the cache was (incorrectly) being used.
3220 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3221 only the Cygwin environment.
3223 11. When return_path_on_delivery was set as a log selector, if different remote
3224 addresses in the same message used different return paths and parallel
3225 remote delivery occurred, the wrong values would sometimes be logged.
3226 (Whenever a remote delivery process finished, the return path value from
3227 the most recently started remote delivery process was logged.)
3229 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3230 lines when AUTH and/or TLS are in use. This is the "received protocol"
3231 field. Exim used to use "asmtp" for authenticated SMTP, without any
3232 indication (in the protocol name) for TLS use. Now it follows the RFC and
3233 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3234 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3235 names appear in log lines as well as in Received: header lines.
3237 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3238 extensions, and comparison tests, and to fix some bugs.
3240 14. Changes to the "personal" filter test:
3242 (1) The test was buggy in that it was just doing the equivalent of
3243 "contains" tests on header lines. For example, if a user's address was
3244 anne@some.where, the "personal" test would incorrectly be true for
3246 To: susanne@some.where
3248 This test is now done by extracting each address from the header in turn,
3249 and checking the entire address. Other tests that are part of "personal"
3250 are now done using regular expressions (for example, to check local parts
3251 of addresses in From: header lines).
3253 (2) The list of non-personal local parts in From: addresses has been
3254 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3255 taken from the Sieve specification recommendations.
3257 (3) If the message contains any header line starting with "List-" it is
3258 treated as non-personal.
3260 (4) The test for "circular" in the Subject: header line has been removed
3261 because it now seems ill-conceived.
3263 15. Minor typos in src/EDITME comments corrected.
3265 16. Installed latest exipick from John Jetmore.
3267 17. If headers_add on a router specified a text string that was too long for
3268 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3269 of string_sprintf() is now avoided.
3271 18. $message_body_size was not set (it was always zero) when running the DATA
3272 ACL and the local_scan() function.
3274 19. For the "mail" command in an Exim filter, no default was being set for
3275 the once_repeat time, causing a random time value to be used if "once" was
3276 specified. (If the value happened to be <= 0, no repeat happened.) The
3277 default is now 0s, meaning "never repeat". The "vacation" command was OK
3278 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3279 (I found it when inspecting the code).
3281 20. There is now an overall timeout for performing a callout verification. It
3282 defaults to 4 times the callout timeout, which applies to individual SMTP
3283 commands during the callout. The overall timeout applies when there is more
3284 than one host that can be tried. The timeout is checked before trying the
3285 next host. This prevents very long delays if there are a large number of
3286 hosts and all are timing out (e.g. when the network connections are timing
3287 out). The value of the overall timeout can be changed by specifying an
3288 additional sub-option for "callout", called "maxwait". For example:
3290 verify = sender/callout=5s,maxwait=20s
3292 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3293 to the end before writing, but this should make it even safer).
3295 22. Exim was forgetting that it had advertised PIPELINING for the second and
3296 subsequent messages on an SMTP connection. It was also not resetting its
3297 memory on STARTTLS and an internal HELO.
3299 23. When Exim logs an SMTP synchronization error within a session, it now
3300 records whether PIPELINING has been advertised or not.
3302 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3303 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3304 rather than long int.
3306 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3308 26. Added the never_mail option to autoreply.
3314 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3315 crash if the getsockname() call failed; this can happen if a connection is
3316 closed very soon after it is established. The problem was simply in the
3317 order in which certain operations were done, causing Exim to try to write
3318 to the SMTP stream before it had set up the file descriptor. The bug has
3319 been fixed by making things happen in the correct order.
3325 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3326 before the connection was closed, thus losing the rejection response.
3328 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3329 some early Solaris releases, but causes trouble in current releases where
3330 socklen_t is defined.
3332 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3335 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3338 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3339 that releasing the top of it at the end releases what was used for sub-
3340 expansions (unless the block got too big). However, discard this block if
3341 the first thing is a variable or header, so that we can use its block when
3342 it is dynamic (useful for very large $message_headers, for example).
3344 6. Lookups now cache *every* query, not just the most recent. A new, separate
3345 store pool is used for this. It can be recovered when all lookup caches are
3346 flushed. Lookups now release memory at the end of their result strings.
3347 This has involved some general refactoring of the lookup sources.
3349 7. Some code has been added to the store_xxx() functions to reduce the amount
3350 of flapping under certain conditions.
3352 8. log_incoming_interface used to affect only the <= reception log lines. Now
3353 it causes the local interface and port to be added to several more SMTP log
3354 lines, for example "SMTP connection from", and rejection lines.
3356 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3358 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3360 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3361 could be overwritten at the end of the current message (or the start of a
3362 new message if it was set in a HELO ACL). The value is now preserved for
3363 the duration of the SMTP connection.
3365 12. If a transport had a headers_rewrite setting, and a matching header line
3366 contained an unqualified address, that address was qualified, even if it
3367 did not match any rewriting rules. The underlying bug was that the values
3368 of the flags that permit the existence of unqualified sender and recipient
3369 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3370 non-local messages, and by -bnq for local messages) were not being
3371 preserved with the message after it was received.
3373 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3374 "next input=" as part of the text comprising the host identity instead of
3375 the correct text. The code was using the same buffer for two different
3376 strings. However, depending on which order the printing function evaluated
3377 its arguments, the bug did not always show up. Under Linux, for example, my
3378 test suite worked just fine.
3380 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3381 doesn't work with some older versions of Perl. It has been changed to "my",
3382 which in any case is probably the better facility to use.
3384 15. A really picky compiler found some instances of statements for creating
3385 error messages that either had too many or two few arguments for the format
3388 16. The size of the buffer for calls to the DNS resolver has been increased
3389 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3390 for addresses that have a lot of PTR records. This alleviates a problem; it
3391 does not fully solve it.
3393 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3394 buffer now truncates the list and logs the incident, which is the same
3395 action as happens when Exim is looking up a host name and its aliases.
3396 Previously in this situation something unpredictable would happen;
3397 sometimes it was "internal error: store_reset failed".
3399 18. If a server dropped the connection unexpectedly when an Exim client was
3400 using GnuTLS and trying to read a response, the client delivery process
3401 crashed while trying to generate an error log message.
3403 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3404 string, for example:
3406 warn message = H1: something\nH2: something
3408 the text was added as a single header line from Exim's point of view
3409 though it ended up OK in the delivered message. However, searching for the
3410 second and subsequent header lines using $h_h2: did not work. This has been
3411 fixed. Similarly, if a system filter added multiple headers in this way,
3412 the routers could not see them.
3414 20. Expanded the error message when iplsearch is called with an invalid key to
3415 suggest using net-iplsearch in a host list.
3417 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3418 ACLs are no longer actually imposed (and a message to that effect is
3421 22. If a "gecos" field in a passwd entry contained escaped characters, in
3422 particular, if it contained a \" sequence, Exim got it wrong when building
3423 a From: or a Sender: header from that name. A second bug also caused
3424 incorrect handling when an unquoted " was present following a character
3425 that needed quoting.
3427 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3428 was not being matched caselessly.
3430 24. Arranged for all hyphens in the exim.8 source to be escaped with
3433 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3434 properly. Recipient callout cache records were still being keyed to include
3435 the sender, even when use_sender was set false. This led to far more
3436 callouts that were necessary. The sender is no longer included in the key
3437 when use_sender is false.
3439 26. Added "control = submission" modifier to ACLs.
3441 27. Added the ${base62d: operator to decode base 62 numbers.
3443 28. dnsdb lookups can now access SRV records.
3445 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3446 the configuration file.
3448 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3449 (-v) mode. This makes the output for a verbose queue run more intelligible.
3451 31. Added a use_postmaster feature to recipient callouts.
3453 32. Added the $body_zerocount variable, containing the number of binary zero
3454 bytes in the message body.
3456 33. The time of last modification of the "new" subdirectory is now used as the
3457 "mailbox time last read" when there is a quota error for a maildir
3460 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3462 35. Added +ignore_unknown as a special item in host lists.
3464 36. Code for decoding IPv6 addresses in host lists is now included, even if
3465 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3466 address was recognized as an IP address, but was then not correctly decoded
3467 into binary, causing unexpected and incorrect effects when compared with
3474 1. Very minor rewording of debugging text in manualroute to say "list of
3475 hosts" instead of "hostlist".
3477 2. If verify=header_syntax was set, and a header line with an unqualified
3478 address (no domain) and a large number of spaces between the end of the
3479 name and the colon was received, the reception process suffered a buffer
3480 overflow, and (when I tested it) crashed. This was caused by some obsolete
3481 code that should have been removed. The fix is to remove it!
3483 3. When running in the test harness, delay a bit after writing a bounce
3484 message to get a bit more predictability in the log output.
3486 4. Added a call to search_tidyup() just before forking a reception process. In
3487 theory, someone could use a lookup in the expansion of smtp_accept_max_
3488 per_host which, without the tidyup, could leave open a database connection.
3490 5. Added the variables $recipient_data and $sender_data which get set from a
3491 lookup success in an ACL "recipients" or "senders" condition, or a router
3492 "senders" option, similar to $domain_data and $local_part_data.
3494 6. Moved the writing of debug_print from before to after the "senders" test
3497 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3498 problems for message scanning, either using a data ACL, or using
3499 local_scan() because the Received: header was not generated till after they
3500 were called (in order to set the time as the time of reception completion).
3501 I have revised the way this works. The header is now generated after the
3502 body is received, but before the ACL or local_scan() are called. After they
3503 are run, the timestamp in the header is updated.
3509 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3510 before starting a queue runner without re-exec. This happened only when
3511 deliver_drop_privilege was set or when the Exim user was set to root. The
3512 effect of the bug was that timeouts during subsequent deliveries caused
3513 crashes instead of being properly handled. The handler is now left at its
3514 default (and expected) setting.
3516 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3517 message, again when deliver_drop_privilege is set or Exim is run as root.
3518 The bug described in (1) was not present in this case, but the tidying up
3519 of the other signals was missing. I have made the two cases consistent.
3521 3. The ignore_target_hosts setting on a manualroute router was being ignored
3522 for hosts that were looked up using the /MX notation.
3524 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3527 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3528 operated on the sender address. After changing the $sender_address to <>
3529 for the sender address verify, Exim was re-instated it as the original
3530 (before rewriting) address, but remembering that it had rewritten it, so it
3531 wasn't rewriting it again. This bug also had the effect of breaking the
3532 sender address verification caching when the sender address was rewritten.
3534 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3535 This has been changed so that if the ip literal address matches
3536 ignore_target_hosts, the router declines.
3538 7. Added expansion conditions match_domain, match_address, and match_local_
3539 part (NOT match_host).
3541 8. The placeholder for the Received: header didn't have a length field set.
3543 9. Added code to Exim itself and to exim_lock to test for a specific race
3544 condition that could lead to file corruption when using MBX delivery. The
3545 issue is with the lockfile that is created in /tmp. If this file is removed
3546 after a process has opened it but before that process has acquired a lock,
3547 there is the potential for a second process to recreate the file and also
3548 acquire a lock. This could lead to two Exim processes writing to the file
3549 at the same time. The added code performs the same test as UW imapd; it
3550 checks after acquiring the lock that its file descriptor still refers to
3551 the same named file.
3553 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3554 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3557 11. Added the smtp_active_hostname option. If used, this will typically be made
3558 to depend on the incoming interface address. Because $interface_address is
3559 not set up until the daemon has forked a reception process, error responses
3560 that can happen earlier (such as "too many connections") no longer contain
3563 12. If an expansion in a condition on a "warn" statement fails because a lookup
3564 defers, the "warn" statement is abandoned, and the next ACL statement is
3565 processed. Previously this caused the whole ACL to be aborted.
3567 13. Added the iplsearch lookup type.
3569 14. Added ident_timeout as a log selector.
3571 15. Added tls_certificate_verified as a log selector.
3573 16. Added a global option tls_require_ciphers (compare the smtp transport
3574 option of the same name). This controls incoming TLS connections.
3576 17. I finally figured out how to make tls_require_ciphers do a similar thing
3577 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3578 before starting the TLS session.
3580 18. Tabs are now shown as \t in -bP output.
3582 19. If the log selector return_path_on_delivery was set, Exim crashed when
3583 bouncing a message because it had too many Received: header lines.
3585 20. If two routers both had headers_remove settings, and the first one included
3586 a superfluous trailing colon, the final name in the first list and the
3587 first name in the second list were incorrectly joined into one item (with a
3588 colon in the middle).
3594 1. Added -C and -D options to the exinext utility, mainly to make it easier
3595 to include in the automated testing, but these could be helpful when
3596 multiple configurations are in use.
3598 2. The exinext utility was not formatting the output nicely when there was
3599 an alternate port involved in the retry record key, nor when there was a
3600 message id as well (for retries that were specific to a specific message
3601 and a specific host). It was also confused by IPv6 addresses, because of
3602 the additional colons they contain. I have fixed the IPv4 problem, and
3603 patched it up to do a reasonable job for IPv6.
3605 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3606 delivery, the log line now contains "pipelined" if PIPELINING was used.
3608 4. An SMTP transport process used to panic and die if the bind() call to set
3609 an explicit outgoing interface failed. This has been changed; it is now
3610 treated in the same way as a connect() failure.
3612 5. A reference to $sender_host_name in the part of a conditional expansion
3613 that was being skipped was still causing a DNS lookup. This no longer
3616 6. The def: expansion condition was not recognizing references to header lines
3617 that used bh_ and bheader_.
3619 7. Added the _cache feature to named lists.
3621 8. The code for checking quota_filecount in the appendfile transport was
3622 allowing one more file than it should have been.
3624 9. For compatibility with Sendmail, the command line option
3632 and sets the incoming protocol and host name (for trusted callers). The
3633 host name and its colon can be omitted when only the protocol is to be set.
3634 Note the Exim already has two private options, -pd and -ps, that refer to
3635 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3636 "s", but I don't think that's a major issue.
3638 10. A number of refactoring changes to the code, none of which should affect
3641 (a) The number of logging options was getting close to filling up the
3642 32-bit word that was used as a bit map. I have split them into two classes:
3643 those that are passed in the argument to log_write(), and those that are
3644 only ever tested independently outside of that function. These are now in
3645 separate 32-bit words, so there is plenty of room for expansion again.
3646 There is no change in the user interface or the logging behaviour.
3648 (b) When building, for example, log lines, the code previously used a
3649 macro that called string_cat() twice, in order to add two strings. This is
3650 not really sufficiently general. Furthermore, there was one instance where
3651 it was actually wrong because one of the argument was used twice, and in
3652 one call a function was used. (As it happened, calling the function twice
3653 did not affect the overall behaviour.) The macro has been replaced by a
3654 function that can join an arbitrary number of extra strings onto a growing
3657 (c) The code for expansion conditions now uses a table and a binary chop
3658 instead of a serial search (which was left over from when there were very
3659 few conditions). Also, it now recognizes conditions like "pam" even when
3660 the relevant support is not compiled in: a suitably worded error message is
3661 given if an attempt is made to use such a condition.
3663 11. Added ${time_interval:xxxxx}.
3665 12. A bug was causing one of the ddress fields not to be passed back correctly
3666 from remote delivery subprocesses. The field in question was not being
3667 subsequently used, so this caused to problems in practice.
3669 13. Added new log selectors queue_time and deliver_time.
3671 14. Might have fixed a bug in maildirsizefile handling that threw up
3672 "unexpected character" debug warnings, and recalculated the data
3673 unnecessarily. In any case, I expanded the warning message to give more
3676 15. Added the message "Restricted characters in address" to the statements in
3677 the default ACL that block characters like @ and % in local parts.
3679 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3680 Three changes have been made:
3682 (a) There was a serious bug; a negative response to MAIL caused the whole
3683 recipient domain to be cached as invalid, thereby blocking all messages
3684 to all local parts at the same domain, from all senders. This bug has
3685 been fixed. The domain is no longer cached after a negative response to
3686 MAIL if the sender used is not empty.
3688 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3691 (c) A new callout option, "use_sender" has been added for people who want
3692 the modified behaviour.
3698 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3701 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3702 indeed breaks things for older releases.
3704 3. Added additional logging to the case where there is a problem reading data
3705 from a filter that is running in a subprocess using a pipe, in order to
3706 try to track down a specific problem.
3708 4. Testing facility fudge: when running in the test harness and attempting
3709 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3710 getting "No route to host". Convert this to a timeout.
3712 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3715 6. Some OS don't have socklen_t but use size_t instead. This affects the
3716 fifth argument of getsockopt() amongst other things. This is now
3717 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3718 can be set for individual OS. I have set it for SunOS5, OSF1, and
3719 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3720 some earlier ones do not.
3722 7. Change 4.30/15 was not doing the test caselessly.
3724 8. The standard form for an IPv6 address literal was being rejected by address
3725 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3726 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3727 this, as well as the form without the "IPv6" on the front (but only when
3728 address literals are enabled, of course).
3730 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3732 10. Exim crashed if a message with an empty sender address specified by -f
3733 encountered a router with an errors_to setting. This could be provoked only
3734 by a command such as
3738 where an empty string was supplied; "<>" did not hit this bug.
3740 11. Installed PCRE release 4.5.
3742 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3743 remained set. It is now erased.
3745 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3746 times from message ids (which are base 36 rather than the normal 62).
3748 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3749 were being counted as actual protocol errors, and logged if the log
3750 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3751 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3752 and DATA following a set of rejected RCPTs do not count as protocol errors.
3753 In other words, Exim assumes they were pipelined, though this may not
3754 actually be the case. Of course, in all cases the client gets an
3755 appropriate error code.
3757 15. If a lookup fails in an ACL condition, a message about the failure may
3758 be available; it is used if testing the ACL cannot continue, because most
3759 such messages specify what the cause of the deferral is. However, some
3760 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3761 that caused an old message to be retained and used if a later statement
3762 caused a defer, replacing the real cause of the deferral.
3764 16. If an IP address had so many PTR records that the DNS lookup buffer
3765 was not large enough to hold them, Exim could crash while trying to process
3766 the truncated data. It now detects and logs this case.
3768 17. Further to 4.21/58, another change has been made: if (and only if) the
3769 first line of a message (the first header line) ends with CRLF, a bare LF
3770 in a subsequent header line has a space inserted after it, so as not to
3771 terminate the header.
3773 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3774 unnecessarily, used atoi() instead of strtol(), and didn't check the
3775 termination when getting file sizes from file names by regex.
3777 19. Completely re-implemented the support for maildirsize files, in the light
3778 of a number of problems with the previous contributed implementation
3779 (4.30/29). In particular:
3781 . If the quota is zero, the maildirsize file is maintained, but no quota is
3784 . If the maildir directory does not exist, it is created before any attempt
3785 to write a maildirsize file.
3787 . The quota value in the file is just a cache; if the quota is changed in
3788 the transport, the new value overrides.
3790 . A regular expression is available for excluding directories from the
3793 20. The autoreply transport checks the characters in options that define the
3794 message's headers; it allows continued headers, but it was checking with
3795 isspace() after an embedded newline instead of explicitly looking for a
3798 21. If all the "regular" hosts to which an address was routed had passed their
3799 expiry times, and had not reached their retry times, the address was
3800 bounced, even if fallback hosts were defined. Now Exim should go on to try
3803 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3804 equivalent code in the SMTP transport. Some hosts send humungous responses
3805 to HELO/EHLO, more than 1024 it seems.
3807 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3808 gives compiler warnings in some environments. I've now done it "properly",
3811 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3812 (because of problems with the built-in one) was declared to return uschar *
3813 instead of char *, causing compiler failure.
3815 25. Fixed a file descriptor leak when processing alias/forward files.
3817 26. Fixed a minor format string issue in dbfn.c.
3819 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3821 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3822 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3825 29. When a sender address is verified, it is cached, to save repeating the test
3826 when there is more than one recipient in a message. However, when the
3827 verification involves a callout, it is possible for different callout
3828 options to be set for different recipients. It is too complicated to keep
3829 track of this in the cache, so now Exim always runs a verification when a
3830 callout is required, relying on the callout cache for the optimization.
3831 The overhead is duplication of the address routing, but this should not be
3834 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3835 to be verified with callout=postmaster, and the main callout worked but the
3836 postmaster check failed, the verification correctly failed. However, if a
3837 subsequent RCPT command asked for sender verification *without* the
3838 postmaster check, incorrect caching caused this verification also to fail,
3841 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3842 it was not caching the DNS options (qualify_single, search_parents) that
3843 were used when the lookup failed. A subsequent lookup with different
3844 options therefore always gave the same answer, though there were cases
3845 where it should not have. (Example: a "domains = !$mx_any" option on a
3846 dnslookup router: the "domains" option is always processed without any
3847 widening, but the router might have qualify_single set.) Now Exim uses the
3848 cached value only when the same options are set.
3850 32. Added John Jetmore's "exipick" utility to the distribution.
3852 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3853 than a timeout (e.g. a certificate is required, and is not provided), an
3854 Exim server now closes the connection immediately. Previously it waited for
3855 the client to close - but if the client is SSL, it seems that they each
3856 wait for each other, leading to a delay before one of them times out.
3858 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3859 maintained 0.8.x compatibility because I don't think many are using it, and
3860 it is clearly obsolete.
3862 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3865 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3866 client certificate was expired. A simple patch fixes this, though I don't
3867 understand the full logic of why the verify callback is called multiple
3870 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3871 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3872 which causes problems with some clients (such as the Certicom SSL Plus
3873 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3874 disables the coutermeasure allowing Eudora to connect."
3876 38. Exim was not checking that a write() to a log file succeeded. This could
3877 lead to Bad Things if a log got too big, in particular if it hit a file
3878 size limit. Exim now panics and dies if it cannot write to a log file, just
3879 as it does if it cannot open a log file.
3881 39. Modified OS/Makefile-Linux so that it now contains
3883 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3885 The two -D definitions ensure that Exim is compiled with large file
3886 support, which makes it possible to handle log files that are bigger than
3889 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3890 instance) a domain was checked against a named list that involved a lookup,
3891 causing $domain_data to be set, then another domain was checked against the
3892 same list, then the first domain was re-checked, the value of $domain_data
3893 after the final check could be wrong. In particular, if the second check
3894 failed, it could be set empty. This bug probably also applied to
3897 41. The strip_trailing_dot option was not being applied to the address given
3898 with the -f command-line option.
3900 42. The code for reading a message's header from the spool was incrementing
3901 $received_count, but never initializing it. This meant that the value was
3902 incorrect (doubled) while delivering a message in the same process in which
3903 it was received. In the most common configuration of Exim, this never
3904 happens - a fresh exec is done - but it can happen when
3905 deliver_drop_privilege is set.
3907 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3908 it now includes up to 150 characters of the unexpected data in the log
3911 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3912 and building data strings. The size of both of these buffers was 10 000
3913 bytes - far larger than anybody would *ever* want, thought I. Needless to
3914 say, somebody hit the limit. I have increased the maximum line length to
3915 20 000 and the maximum data length of concatenated lines to 100 000. I have
3916 also fixed two bugs, because there was no checking on these buffers. Tsk,
3917 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3918 buffer is too small.
3920 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3921 lsearch lookups. Now it does.
3923 46. When parsing a route_list item in a manualroute router, a fixed-length
3924 buffer was used for the list of hosts. I made this 1024 bytes long,
3925 thinking that nobody would ever have a list of hosts that long. Wrong.
3926 Somebody had a whole pile of complicated expansion conditions, and the
3927 string was silently truncated, leading to an expansion error. It turns out
3928 that it is easier to change to an unlimited length (owing to other changes
3929 that have happened since this code was originally written) than to build
3930 structure for giving a limitation error. The length of the item that
3931 expands into the list of hosts is now unlimited.
3933 47. The lsearch lookup could not handle data where the length of text line was
3934 more than 4095 characters. Such lines were truncated, leading to shortened
3935 data being returned. It should now handle lines of any length.
3937 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3938 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3941 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3942 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3943 no longer happens in comment lines. A list of replacements is now placed
3944 at the head of all of the source files, except those whose only change is
3945 to replace PERL_COMMAND in the very first #! line.
3947 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3948 messages on the queue, with a bottom-up merge sort, using code contributed
3949 by Michael Haardt. This should make operations like -bp somewhat faster on
3950 large queues. It won't affect queue runners, except when queue_run_in_order
3953 51. Installed eximstats 1.31 in the distribution.
3955 52. Added support for SRV lookups to the dnslookup router.
3957 53. If an ACL referred to $message_body or $message_body_end, the value was not
3958 reset for any messages that followed in the same SMTP session.
3960 54. The store-handling optimization for building very long strings was not
3961 differentiating between the different store pools. I don't think this
3962 actually made any difference in practice, but I've tidied it.
3964 55. While running the routers to verify a sender address, $sender_address
3965 was still set to the sender address. This is wrong, because when routing to
3966 send a bounce to the sender, it would be empty. Therefore, I have changed
3967 it so that, while verifying a sender address, $sender_address is set to <>.
3968 (There is no change to what happens when verifying a recipient address.)
3970 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3971 target A or AAAA records (if not already returned) without resetting the
3972 qualify_single or search_parents options of the DNS resolver. These are
3973 inappropriate in this case because the targets of MX and SRV records must
3974 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3975 target that, when qualified, matched something in the local domain. These
3976 two options are now turned off when doing these lookups.
3978 57. It seems that at least some releases of Reiserfs (which does not have the
3979 concept of a fixed number of inodes) returns zero and not -1 for the
3980 number of available inodes. This interacted badly with check_spool_inodes,
3981 which assumed that -1 was the "no such thing" setting. What I have done is
3982 to check that the total number of inodes is greater than zero before doing
3983 the test of how many are available.
3985 58. When a "warn" ACL statement has a log_message modifier, the message is
3986 remembered, and not repeated. This is to avoid a lot of repetition when a
3987 message has many recipients that cause the same warning to be written.
3988 Howewer, Exim was preserving the list of already written lines for an
3989 entire SMTP session, which doesn't seem right. The memory is now reset if a
3990 new message is started.
3992 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3993 debugging output unless log_rewrite was also set.
3995 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3996 of (int)(handle) when we know that handle contains (void *)(-1).
3998 61. The Exim daemon panic-logs an error return when it closes the incoming
3999 connection. However "connection reset by peer" seems to be common, and
4000 isn't really an error worthy of noting specially, so that particular error
4003 62. When Exim is trying to find all the local interfaces, it used to panic and
4004 die if the ioctl to get the interface flags failed. However, it seems that
4005 on at least one OS (Solaris 9) it is possible to have an interface that is
4006 included in the list of interfaces, but for which you get a failure error
4007 for this call. This happens when the interface is not "plumbed" into a
4008 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4009 failure of the "get flags" call assumes that the interface is down.
4011 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4012 makes life easier for people who are doing arithmetic on fields extracted
4013 from dates, where you often get leading zeros that should not be
4014 interpreted as octal.
4016 64. Added qualify_domain to the redirect router, to override the global
4019 65. If a pathologically long header line contained very many addresses (the
4020 report of this problem mentioned 10 000) and each of them was rewritten,
4021 Exim could use up a very large amount of memory. (It kept on making new
4022 copies of the header line as it rewrote, and never released the old ones.)
4023 At the expense of a bit more processing, the header rewriting function has
4024 been changed so that it no longer eats memory in this way.
4026 66. The generation of the Received: header has been moved from the time that a
4027 message starts to be received, to the time that it finishes. The timestamp
4028 in the Received: header should now be very close to that of the <= log
4029 line. There are two side-effects of this change:
4031 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4032 logged header lines no longer include the local Received: line, because
4033 it has not yet been created. The same applies to a copy of the message
4034 that is returned to a non-SMTP sender when a message is rejected.
4036 (b) When a filter file is tested using -bf, no additional Received: header
4037 is added to the test message. After some thought, I decided that this
4040 This change does not affect the value of $received_for. It is still set
4041 after address rewriting, but before local_scan() is called.
4043 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4045 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4046 gave an unhelpful panic error message, and a defer error. I have managed to
4047 change this behaviour so that it now rejects any supplied certificate,
4048 which seems right, as the list of acceptable certificates is empty.
4050 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4051 gave an unhelpful defer error. I have not managed to make this reject any
4052 supplied certificates, but the error message it gives is "no certificate
4053 supplied", which is not helpful.
4055 70. exigrep's output now also includes lines that are not associated with any
4056 message, but which match the given pattern. Implemented by a patch from
4057 Martin Sluka, which also tidied up the Perl a bit.
4059 71. Recipient callout verification, like sender verification, was using <> in
4060 the MAIL FROM command. This isn't really the right thing, since the actual
4061 sender may affect whether the remote host accepts the recipient or not. I
4062 have changed it to use the actual sender in the callout; this means that
4063 the cache record is now keyed on a recipient/sender pair, not just the
4064 recipient address. There doesn't seem to be a real danger of callout loops,
4065 since a callout by the remote host to check the sender would use <>.
4066 [SEE ABOVE: changed after hitting problems.]
4068 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4069 temporary errors. However, in the case of such a code being given after
4070 the end of a data transmission (i.e. after ".") Exim was failing to write
4071 a retry record for the message. (Yes, there was some broken host that was
4072 actually sending 8xx at this point.)
4074 73. An unknown lookup type in a host list could cause Exim to panic-die when
4075 the list was checked. (An example that provoked this was putting <; in the
4076 middle of a list instead of at the start.) If this happened during a DATA
4077 ACL check, a -D file could be left lying around. This kind of configuration
4078 error no longer causes Exim to die; instead it causes a defer errror. The
4079 incident is still logged to the main and panic logs.
4081 74. Buglet left over from Exim 3 conversion. The message "too many messages
4082 in one connection" was written to the rejectlog but not the mainlog, except
4083 when address rewriting (yes!) was being logged.
4085 75. Added write_rejectlog option.
4087 76. When a system filter was run not as root (that is, when system_filter_user
4088 was set), the values of the $n variables were not being returned to the
4089 main process; thus, they were not subsequently available in the $sn
4092 77. Added +return_path_on_delivery log selector.
4094 78. A connection timeout was being treated differently from recipients deferred
4095 when testing hosts_max_try with a message that was older than the host's
4096 retry timeout. (The host should not be counted, thus allowing all hosts to
4097 be tried at least once before bouncing.) This may have been the cause of an
4098 occasionally reported bug whereby a message would remain on the queue
4099 longer than the retry timeout, but would be bounced if a delivery was
4100 forced. I say "may" because I never totally pinned down the problem;
4101 setting up timeout/retry tests is difficult. See also the next item.
4103 79. The ultimate address timeout was not being applied to errors that involved
4104 a combination of host plus message (for example, a timeout on a MAIL
4105 command). When an address resolved to a number of possible hosts, and they
4106 were not all tried for each delivery (e.g. because of hosts_max_try), a
4107 message could remain on the queue longer than the retry timeout.
4109 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4112 81. Fixed an obscure SMTP outgoing bug which required at least the following
4113 conditions: (a) there was another message waiting for the same server;
4114 (b) the server returned 5xx to all RCPT commands in the first message so
4115 that the message was not completed; (c) the server dropped the connection
4116 or gave a negative response to the RSET that Exim sends to abort the
4117 transaction. The observed case was a dropped connection after DATA that had
4118 been sent in pipelining mode. That is, the server had advertised PIPELINING
4119 but was not implementing it correctly. The effect of the bug was incorrect
4120 behaviour, such as trying another host, and this could lead to a crash.
4126 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4127 and daemon.c were passed as pointers to ints; they should have been
4128 pointers to socklen_t variables (which are typically unsigned ints).
4130 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4133 3. Fixed a really odd bug that affected only the testing scheme; patching a
4134 certain fixed string in the binary changed the value of another string that
4135 happened to be identical to the end of the original first string.
4137 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4138 name", it returns that address as the IP address. On some operating
4139 systems (e.g. Solaris), it also passes back the IP address string as the
4140 "host name". However, on others (e.g. Linux), it passes back an empty
4141 string. Exim wasn't checking for this, and was changing the host name to an
4142 empty string, assuming it had been canonicized.
4144 5. Although rare, it is permitted to have more than one PTR record for a given
4145 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4146 all the names associated with an address, because they do in Solaris.
4147 However, it seems that they do not in Linux for data that comes from the
4148 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4149 I found this out when I moved to a new Linux workstation and tried to run
4150 the Exim test suite.
4152 To get round this problem I have changed the code so that it now does its
4153 own call to the DNS to look up PTR records when searching for a host name.
4154 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4155 addresses that are only in /etc/hosts are still found.
4157 This behaviour is, however, controlled by an option called host_lookup_
4158 order, which defaults to "bydns:byaddr". If people want to use the other
4159 order, or indeed, just use one or the other means of lookup, they can
4160 specify it in this variable.
4162 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4163 some operating systems, this comes back from gethostbyaddr() as an empty
4164 string, and this is what Exim used to test for. However, it seems that in
4165 other systems, "." is yielded. Exim now tests for this case too.
4167 7. The values of check_spool_space and check_log_space are now held internally
4168 as a number of kilobytes instead of an absolute number of bytes. If a
4169 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4170 kilobyte. This means that much larger values can be stored.
4172 8. Exim monitor: an attempt to get the action menu when not actually pointing
4173 at a message produces an empty menu entitled "No message selected". This
4174 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4175 no entries in it ("Shell widget menu has zero width and/or height"). So I
4176 have added a single, blank menu entry in this case.
4178 9. Added ${quote_local_part.
4180 10. MIME decoding is now applied to the contents of Subject: header lines when
4183 11. Now that a reference to $sender_host_address automatically causes a reverse
4184 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4185 host lookup before query-style lookups in lists that might use this
4186 variable. This has therefore been abolished, and the "net-" prefix is no
4187 longer necessary for query-style lookups.
4189 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4190 have been a typo for LFLAGS, so it has been changed.
4192 13. The install script calls Exim with "-C /dev/null" in order to find the
4193 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4194 to be output. Howeve, since Exim outputs its version number before the
4195 error, it didn't break the script. It just looked ugly. I fixed this by
4196 always allowing "-C /dev/null" if the caller is root.
4198 14. Ignore overlarge ACL variable number when reading spool file - insurance
4199 against a later release with more variables having written the file.
4201 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4202 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4203 this, as well as the form without the "IPv6" on the front.
4205 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4206 OS/Makefile-Darwin file.
4208 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4209 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4211 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4212 variable. However, some OS use ints for inodes. I've added cast to long int
4213 to get rid of the compiler warning.
4215 19. I had forgotten to lock out "/../" in configuration file names when
4216 ALT_CONFIG_PREFIX was set.
4218 20. Routers used for verification do not need to specify transports. However,
4219 if such a router generated a host list, and callout was configured, Exim
4220 crashed, because it could not find a port number from the (non-existent)
4221 transport. It now assumes port 25 in this circumstance.
4223 21. Added the -t option to exigrep.
4225 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4226 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4227 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4228 them is set, LOOKUP_LSEARCH is forced.
4230 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4232 24. Added sender and host information to the "rejected by local_scan()" log
4233 line; previously there was no indication of these.
4235 25. Added .include_if_exists.
4237 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4238 Linux. It turns out that not all file systems support this. Apparently some
4239 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4240 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4241 not supported on the file descriptor, is now ignored when Exim is trying to
4242 sync a directory. This applies only to Linux.
4244 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4246 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4247 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4250 29. Added support for maildirsize files from supplied patch (modified a bit).
4252 30. The use of :fail: followed by an empty string could lead Exim to respond to
4253 sender verification failures with (e.g.):
4255 550 Verification failed for <xxx>
4256 550 Sender verify failed
4258 where the first response line was missing the '-' that indicates it is not
4259 the final line of the response.
4261 31. The loop for finding the name of the user that called Exim had a hardwired
4262 limit of 10; it now uses the value of finduser_retries, which is used for
4263 all other user lookups.
4265 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4268 33. Exim was neglecting to zero errno before one call of strtol() when
4269 expanding a string and expecting an integer value. On some systems this
4270 resulted in spurious "integer overflow" errors. Also, it was casting the
4271 result into an int without checking.
4273 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4274 did not work. The code looks as if it has *never* worked, though it appears
4275 to have been documented since at least releast 1.62. I have made it work.
4277 35. The "timeout_DNS" error in retry rules, also documented since at least
4278 1.62, also never worked. As it isn't clear exactly what this means, and
4279 clearly it isn't a major issue, I have abolished the feature by treating it
4280 as "timeout", and writing a warning to the main and panic logs.
4282 36. The display of retry rules for -brt wasn't always showing the error code
4285 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4286 timeout_connect_A, timeout_connect_MX.
4288 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4289 to the empty sender.
4291 39. The daemon was not analysing the content of -oX till after it had closed
4292 stderr and disconnected from the controlling terminal. This meant that any
4293 syntax errors were only noted on the panic log, and the return code from
4294 the command was 0. By re-arranging the code a little, I've made the
4295 decoding happen first, so such errors now appear on stderr, and the return
4296 code is 1. However, the actual setting up of the sockets still happens in
4297 the disconnected process, so errors there are still only recorded on the
4300 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4301 connections (as happens on some IP stacks) was logged at start up time as
4302 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4303 it from "IPv6 and IPv4", which means that two separate sockets are being
4306 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4307 says whether AF_INET or AF_INET6 was passed as an argument.
4309 42. Exiwhat output was messed up when time zones were included in log
4312 43. Exiwhat now gives more information about the daemon's listening ports,
4313 and whether -tls-on-connect was used.
4315 44. The "port" option of the smtp transport is now expanded.
4317 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4318 silently ignored. Now an error message is written to the main and panic
4321 46. There's a new ACL modifier called "logwrite" which writes to a log file
4322 as soon as it is encountered.
4324 47. Added $local_user_uid and $local_user_gid at routing time.
4326 48. Exim crashed when trying to verify a sender address that was being
4329 49. Exim was recognizing only a space character after ".include". It now also
4330 recognizes a tab character.
4332 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4333 extracting the relevant information from the specification. The man page no
4334 longer contains scrambled data for the -d option, and I've added a section
4335 at the front about calling Exim under different names.
4337 51. Added "extra_headers" argument to the "mail" command in filter files.
4339 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4342 53. Installed eximstats 1.29.
4344 54. Added transport_filter_timeout as a generic transport option.
4346 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4347 Cc: header lines. This was required by RFC 822, but it not required by RFC
4350 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4351 incoming messages that did not have them. Now it does so only if the
4352 message originates locally, that is, if there is no associated remote host
4353 address. When Resent- header lines are present, this applies to the Resent-
4354 lines rather than the non-Resent- lines.
4356 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4357 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4359 58. Messages for configuration errors now include the name of the main
4360 configuration file - useful now that there may be more than one file in a
4361 list (.included file names were always shown).
4363 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4364 for those rare installations that do not start the daemon as root or run it
4365 setuid root. I've cut out the call to initgroups() if the daemon is not
4368 60. The Exim user and group can now be bound into the binary as text strings
4369 that are looked up at the start of Exim's processing.
4371 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4373 62. Added $mailstore_basename variable.
4375 63. Installed patch to sieve.c from Michael Haardt.
4377 64. When Exim failed to open the panic log after failing to open the main log,
4378 the original message it was trying to log was written to stderr and debug
4379 output, but if they were not available (the usual case in production), it
4380 was lost. Now it is written to syslog before the two lines that record the
4381 failures to open the logs.
4383 65. Users' Exim filters run in subprocesses under the user's uid. It is
4384 possible for a "deliver" command or an alias in a "personal" command to
4385 provoke an address rewrite. If logging of address rewriting is configured,
4386 this fails because the process is not running as root or exim. There may be
4387 a better way of dealing with this, but for the moment (because 4.30 needs
4388 to be released), I have disabled address rewrite logging when running a
4389 filter in a non-root, non-exim process.
4395 1. The buildconfig auxiliary program wasn't quoting the value set for
4396 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4397 not defined. This bug was masked in 4.22 by the effect that was fixed in
4400 2. Some messages that were rejected after a message id was allocated were
4401 shown as "incomplete" by exigrep. It no longer does this for messages that
4402 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4404 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4405 have allow_domain_literals set, the ID did not get logged in the <= line.
4406 Domain literals are now always recognized in Message-ID: header lines.
4408 4. The first argument for a ${extract expansion item is the key name or field
4409 number. Leading and trailing spaces in this item were not being ignored,
4410 causing some misleading effects.
4412 5. When deliver_drop_privilege was set, single queue runner processes started
4413 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4414 same command in the process it spins off) were not dropping privilege.
4416 6. When the daemon running as "exim" started a queue runner, it always
4417 re-executed Exim in the spun-off process. This is a waste of effort when
4418 deliver_drop_privilege is set. The new process now just calls the
4419 queue-runner function directly.
4425 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4428 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4429 ignored. Though the use of -oP was forcing the writing of a pid file, it
4430 was always written to the default place.
4432 3. If the message "no IP address found for host xxxx" is generated during
4433 incoming verification, it is now followed by identification of the incoming
4434 connection (so you can more easily find what provoked it).
4436 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4438 5. Added some features to "harden" Exim a bit more against certain attacks:
4440 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4441 be put in Local/Makefile. This is like the never_users runtime option,
4442 but it cannot be overridden. The default setting is "root".
4444 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4445 prefix string with which any file named in a -C command line option
4448 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4449 is retained for -C and -D only if the caller of Exim is root. Without
4450 it, the exim user may also use -C and -D and retain privilege.
4452 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4453 command line option is disabled.
4455 6. Macro names set by the -D option must start with an upper case letter, just
4456 like macro names defined in the configuration file.
4458 7. Added "dereference=" facility to LDAP.
4460 8. Two instances of the typo "uknown" in the source files are fixed.
4462 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4463 the Configure-Makefile script screwed up while processing it.
4465 10. Incorporated PCRE 4.4.
4467 11. The SMTP synchronization check was not operating right at the start of an
4468 SMTP session. For example, it could not catch a HELO sent before the client
4469 waited for the greeting. There is now a check for outstanding input at the
4470 point when the greeting is written. Because of the duplex, asynchronous
4471 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4472 way, but not yet received, when the check is performed.
4474 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4475 on TCP/IP sockets, because this apparently causes some broken clients to
4478 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4479 unchanged) from the Cygwin maintainer.
4481 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4482 was supported instead of testing for mbx. Effectively a typo.
4484 15. The spa authenticator server code was not checking that the input it
4485 received was valid base64.
4487 16. The debug output line for the "set" modifier in ACLs was not showing the
4488 name of the variable that was being set.
4490 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4492 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4493 lookup on its own. Something else was needed to trigger the lookup. For
4494 example, a match in host_lookup or the need for a host name in a host list.
4495 Now, if $sender_host_name is referenced and the host name has not yet been
4496 looked up, a lookup is performed. If the lookup fails, the variable remains
4497 empty, and $host_lookup_failed is set to "1".
4499 19. Added "eqi" as a case-independent comparison operator.
4501 20. The saslauthd authentication condition could segfault if neither service
4502 nor realm was specified.
4504 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4505 error message that was logged was misleading, and incoming SMTP
4506 connections were dropped. The message is now more accurate, and temporary
4507 errors are given to SMTP connections.
4509 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4510 (except RSET) with a 421 error, until QUIT is received. However, it was
4511 failing to send a response to QUIT.
4513 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4514 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4515 ACL is now run after the helo_try_verify_hosts code.
4517 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4518 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4519 case-independent, so other case variants are also recognized). Apparently
4520 some systems use these upper case variants.
4522 25. If more than two messages were waiting for the same host, and a transport
4523 filter was specified for the transport, Exim sent two messages over the
4524 same TCP/IP connection, and then failed with "socket operation on non-
4525 socket" when it tried to send the third.
4527 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4529 27. The extern definition of crypt16() in expand.c was not being excluded when
4530 the OS had its own crypt16() function.
4532 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4533 as a preferred synonym for return_size_limit, both as an option and as an
4536 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4538 30. Changed the default configuration ACL to relax the local part checking rule
4539 for addresses that are not in any local domains. For these addresses,
4540 slashes and pipe symbols are allowed within local parts, but the sequence
4541 /../ is explicitly forbidden.
4543 31. SPA server authentication was not clearing the challenge buffer before
4546 32. log_message in a "warn" ACL statement was writing to the reject log as
4547 well as to the main log, which contradicts the documentation and doesn't
4548 seem right (because no rejection is happening). So I have stopped it.
4550 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4551 However, I am unable to do any testing of this.
4553 34. Fixed an infelicity in the appendfile transport. When checking directories
4554 for a mailbox, to see if any needed to be created, it was accidentally
4555 using path names with one or more superfluous leading slashes; tracing
4556 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4558 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4559 contents are added to the log line that is written for every discarded
4560 recipient. (Previously a log_message setting was ignored.)
4562 36. The ${quote: operator now quotes the string if it is empty.
4564 37. The install script runs exim in order to find its version number. If for
4565 some reason other than non-existence or emptiness, which it checks, it
4566 could not run './exim', it was installing it with an empty version number,
4567 i.e. as "exim-". This error state is now caught, and the installation is
4570 38. An argument was missing from the function that creates an error message
4571 when Exim fails to connect to the socket for saslauthd authentication.
4572 This could cause Exim to crash, or give a corrupted message.
4574 39. Added isip, isip4, and isip6 to ${if conditions.
4576 40. The ACL variables $acl_xx are now saved with the message, and can be
4577 accessed later in routers, transports, and filters.
4579 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4580 strings in the file are not string-expanded.
4582 42. If a MAIL command specified a SIZE value that was too large to fit into an
4583 int variable, the check against message_size_limit failed. Such values are
4584 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4585 day this will have to be increased, but I don't think I want to be around
4586 when emails are that large.
4593 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4594 iconv() is not standard in FreeBSD.
4596 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4597 IPv6 enabled. The observed symptom was a segmentation fault on return from
4598 the function os_common_find_running_interfaces() in src/os.c.
4600 3. In the check_special_case() function in daemon.c I had used "errno" as an
4601 argument name, which causes warnings on some systems. This was basically a
4602 typo, since it was named "eno" in the comments!
4604 4. The code that waits for the clock to tick (at a resolution of some fraction
4605 of a second) so as to ensure message-id uniqueness was always waiting for
4606 at least one whole tick, when it could have waited for less. [This is
4607 almost certainly not relevant at current processor speeds, where it is
4608 unlikely to ever wait at all. But we try to future-proof.]
4610 5. The function that sleeps for a time interval that includes fractions of a
4611 second contained a race. It did not block SIGALRM between setting the
4612 timer, and suspending (a couple of lines later). If the interval was short
4613 and the sigsuspend() was delayed until after it had expired, the suspension
4614 never ended. On busy systems this could lead to processes getting stuck for
4617 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4618 process, before it forks any delivery processes. The open lookup caching
4619 mechanism meant that the open file or database connection was passed into
4620 the delivery process. The problem was that delivery processes always tidy
4621 up cached lookup data. This could cause a problem for the next delivery
4622 process started by the queue runner, because the external queue runner
4623 process does not know about the closure. So the next delivery process
4624 still has data in the lookup cache. In the case of a file lookup, there was
4625 no problem because closing a file descriptor in a subprocess doesn't affect
4626 the parent. However, if the lookup was caching a connection to a database,
4627 the connection was closed, and the second delivery process was likely to
4628 see errors such as "PGSQL: query failed: server closed the connection
4629 unexpectedly". The problem has been fixed by closing all cached lookups
4630 in a queue runner before running a delivery process.
4632 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4633 seem to have the "const" qualifier which it has on other OS. I've
4636 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4637 *of the same type* (client or server) with the same public name that an
4638 error should be diagnosed.
4640 9. When Exim looked up a host name for an IP address, but failed to find the
4641 original IP address when looking up the host name (a safety check), it
4642 output the message "<ip address> does not match any IP for NULL", which was
4643 confusing, to say the least. The bug was that the host name should have
4644 appeared instead of "NULL".
4646 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4647 user that is built into the binary, and the -C or -D options is used, root
4648 privilege is dropped before the configuration file is read. In addition,
4649 logging is switched to stderr instead of the normal log files. If the
4650 configuration then re-defines the Exim user, the unprivileged environment
4651 is probably not what is expected, so Exim logs a panic warning message (but
4654 However, if deliver_drop_privilege is set, the unprivileged state may well
4655 be exactly what is intended, so the warning has been cut out in that case,
4656 and Exim is allowed to try to write to its normal log files.
4662 1. smtp_return_error_details was not giving details for temporary sender
4663 or receiver verification errors.
4665 2. Diagnose a configuration error if two authenticators have the same public
4668 3. Exim used not to create the message log file for a message until the first
4669 delivery attempt. This could be confusing when incoming messages were held
4670 for policy or load reasons. The message log file is now created at the time
4671 the message is received, and an initial "Received" line is written to it.
4673 4. The automatically generated man page for command line options had a minor
4674 bug that caused no ill effects; however, a more serious problem was that
4675 the procedure for building the man page automatically didn't always
4676 operate. Consequently, release 4.20 contains an out-of-date version. This
4677 shouldn't happen again.
4679 5. When building Exim with embedded Perl support, the script that builds the
4680 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4681 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4683 6. The freeze_tell option was not being used for messages that were frozen on
4684 arrival, either by an ACL or by local_scan().
4686 7. Added the smtp_incomplete_transaction log selector.
4688 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4689 was accepting AUTH without a new EHLO.
4691 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4692 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4695 10. Logging of TCP/IP connections (when configured) now happens in the main
4696 daemon process instead of the child process, so that the TCP/IP connection
4697 count is more accurate (but it can never be perfect).
4699 11. The use of "drop" in a nested ACL was not being handled correctly in the
4700 outer ACL. Now, if condition failure induced by the nested "drop" causes
4701 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4702 or "require"), the connection is dropped.
4704 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4705 that yield "discard" can now be used with an "accept" or a "discard" verb,
4706 but an error is generated for any others (because I can't see a useful way
4707 to define what should happen).
4709 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4710 are now processed in the same way as lines in the Exim configuration file.
4711 In particular, continuation lines are supported.
4713 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4715 15. Added -ti meaning -t -i.
4717 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4718 domains, and warn by logging if others are found.
4720 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4721 returned from GIFCONF when Exim is trying to find the list of interfaces on
4722 a host. The code in os.c has been modified to copy each ifreq to an aligned
4723 structure in all cases.
4725 Also, in some cases, the returned ifreq's were being copied to a 'struct
4726 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4727 means the last couple of bytes of an IPv6 address could be chopped if the
4728 ifreq contained only a normal sockaddr (14 bytes storage).
4730 18. Named domain lists were not supported in the hosts_treat_as_local option.
4731 An entry such as +xxxx was not recognized, and was treated as a literal
4734 19. Ensure that header lines added by a DATA ACL are included in the reject log
4735 if the ACL subsequently rejects the message.
4737 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4738 MD5 (which is deprecated).
4740 21. When testing a filter file using -bf, Exim was writing a message when it
4741 took the sender from a "From " line in the message, but it was not doing so
4742 when it took $return_path from a Return-Path: header line. It now does.
4744 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4745 with a valid header line field name (a series of printing characters
4746 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4748 23. Changed "disc" in the source to "disk" to conform to the documentation and
4749 the book and for uniformity.
4751 24. Ignore Sendmail's -Ooption=value command line item.
4753 25. When execve() failed while trying to run a command in a pipe transport,
4754 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4755 could be confused with a return value of 69 from the command itself. This
4756 has been changed to 127, the value the shell returns if it is asked to run
4757 a non-existent command. The wording for the related log line suggests a
4758 non-existent command as the problem.
4760 26. If received_header_text expands to an empty string, do not add a Received:
4761 header line to the message. (Well, it adds a token one on the spool, but
4762 marks it "old" so that it doesn't get used or transmitted.)
4764 27. Installed eximstats 1.28 (addition of -nt option).
4766 28. There was no check for failure on the call to getsockname() in the daemon
4767 code. This can fail if there is a shortage of resources on the system, with
4768 ENOMEM, for example. A temporary error is now given on failure.
4770 29. Contrary to the C standard, it seems that in some environments, the
4771 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4772 program. Exim now does this explicitly; it affects the formatting of
4773 timestamps using strftime().
4775 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4776 complaints. I've now initialized all the variables, to avoid this.
4778 32. Header lines added by a system filter were not being "seen" during
4779 transport-time rewrites.
4781 33. The info_callback() function passed to OpenSSL is set up with type void
4782 (*)(SSL *, int, int), as described somewhere. However, when calling the
4783 function (actually a macro) that sets it up, the type void(*)() is
4784 expected. I've put in a cast to prevent warnings from picky compilers.
4786 34. If a DNS black list lookup found a CNAME record, but there were no A
4787 records associated with the domain it pointed at, Exim crashed.
4789 35. If a DNS black list lookup returned more than one A record, Exim ignored
4790 all but the first. It now scans all returned addresses if a particular IP
4791 value is being sought. In this situation, the contents of the
4792 $dnslist_value variable are a list of all the addresses, separated by a
4795 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4796 to accept a host name and all its aliases if the forward lookup for any of
4797 them yielded the IP address of the incoming connection. Now it accepts only
4798 those names whose forward lookup yields the correct IP address. Any other
4799 names are discarded. This closes a loophole whereby a rogue DNS
4800 administrator could create reverse DNS records to break through a
4801 wildcarded host restriction in an ACL.
4803 37. If a user filter or a system filter that ran in a subprocess used any of
4804 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4805 the wrong values were passed to the pipe command ($thisaddress had the
4806 value of $0, $0 had the value of $1, etc). This bug was introduced by
4807 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4809 38. Improved the line breaking for long SMTP error messages from ACLs.
4810 Previously, if there was no break point between 40 and 75 characters, Exim
4811 left the rest of the message alone. Two changes have been made: (a) I've
4812 reduced the minimum length to 35 characters; (b) if it can't find a break
4813 point between 35 and 75 characters, it looks ahead and uses the first one
4814 that it finds. This may give the occasional overlong line, but at least the
4815 remaining text gets split now.
4817 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4818 file descriptors might be low, and that setting 1000 would always raise it.
4819 It turns out that in some environments, the limit is already over 1000 and
4820 that lowering it causes trouble. So now Exim takes care not to decrease it.
4822 40. When delivering a message, the value of $return_path is set to $sender_
4823 address at the start of routing (routers may change the value). By an
4824 oversight, this default was not being set up when an address was tested by
4825 -bt or -bv, which affected the outcome if any router or filter referred to
4828 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4829 log only when "message" or "log_message" are set. However, if one of the
4830 conditions was an address verification, or a call to a nested ACL, the
4831 messages generated by the underlying test were being passed through. This
4832 no longer happens. The underlying message is available in $acl_verify_
4833 message for both "message" and "log_message" expansions, so it can be
4834 passed through if needed.
4836 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4837 new expansion $bh_ to give the encoded byte string without charset
4838 translation. Translation happens only if iconv() is available; HAVE_ICONV
4839 indicates this at build time. HEADERS_CHARSET gives the charset to
4840 translate to; headers_charset can change it in the configuration, and
4841 "headers charset" can change it in an individual filter file.
4843 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4844 that creates RFC 2047 encoded "words" labels them as that charset instead
4845 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4846 expansion operator; (ii) when Exim creates a From: line for a local
4847 message; (iii) when a header line is rewritten to include a "phrase" part.
4849 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4850 buggy, causing it to skip the first lines of messages whose message ID
4851 ended in 'D'. This would not have bitten before Exim release 4.14, because
4852 message IDs were unlikely to end in 'D' before then. The effect was to have
4853 incorrect size information for certain domains.
4855 45. #include "config.h" was missing at the start of the crypt16.c module. This
4856 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4859 46. If there was a timeout during a "random" callout check, Exim treated it as
4860 a failure of the random address, and carried on sending RSET and the real
4861 address. If the delay was just some slowness somewhere, the response to the
4862 original RCPT would be taken as a response to RSET and so on, causing
4863 mayhem of various kinds.
4865 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4866 when I implemented it. It didn't allow for the fact that some option values
4867 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4868 the right test for positive values.
4870 48. Domain names in DNS records are case-independent. Exim always looks them up
4871 in lower case. Some resolvers return domain names in exactly the case they
4872 appear in the zone file, that is, they may contain uppercase letters. Not
4873 all resolvers do this - some return always lower case. Exim was treating a
4874 change of case by a resolver as a change of domain, similar to a widening
4875 of a domain abbreviation. This triggered its re-routing code and so it was
4876 trying to route what was effectively the same domain again. This normally
4877 caused routing to fail (because the router wouldn't handle the domain
4878 twice). Now Exim checks for this case specially, and just changes the
4879 casing of the domain that it ultimately uses when it transmits the message
4882 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4885 50. If a filter generated a file delivery with a non-absolute name (possible if
4886 no home directory exists for the router), the forbid_file option was not
4889 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4890 the existing equality matching.
4892 52. Exim was using ints instead of ino_t variables in some places where it was
4893 dealing with inode numbers.
4895 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4896 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4897 called TMPDIR, and if it finds it is different, it changes its value.
4899 54. The smtp_printf() function is now made available to local_scan() so
4900 additional output lines can be written before returning. There is also an
4901 smtp_fflush() function to enable the detection of a dropped connection.
4902 The variables smtp_input and smtp_batched_input are exported to
4905 55. Changed the default runtime configuration: the message "Unknown user"
4906 has been removed from the ACL, and instead placed on the localuser router,
4907 using the cannot_route_message feature. This means that any verification
4908 failures that generate their own messages won't get overridden. Similarly,
4909 the "Unrouteable address" message that was in the ACL for unverifiable
4910 relay addresses has also been removed.
4912 56. Added hosts_avoid_esmtp to the smtp transport.
4914 57. The exicyclog script was not checking for the esoteric option
4915 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4916 will work only if exicyclog is run under the appropriate euid.
4918 58. Following a discussion on the list, the rules by which Exim recognises line
4919 endings on incoming messages have been changed. The -dropcr and drop_cr
4920 options are now no-ops, retained only for backwards compatibility. The
4921 following line terminators are recognized: LF CRLF CR. However, special
4922 processing applies to CR:
4924 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4925 nor a local message in the state where . is a terminator.
4927 (ii) If a bare CR is encountered in a header line, an extra space is added
4928 after the line terminator so as not to end the header. The reasoning
4929 behind this is that bare CRs in header lines are most likely either
4930 to be mistakes, or people trying to play silly games.
4932 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4933 was being incorrectly given as 18 bytes larger than it should have been.
4934 This is a VOB (very old bug).
4936 60. This may never have affected anything current, but just in case it has:
4937 When the local host is found other than at the start of a list of hosts,
4938 the local host, those with the same MX, and any that follow, are discarded.
4939 When the list in question was part of a longer list of hosts, the following
4940 hosts (not currently being processed) were also being discarded. This no
4941 longer happens. I'm not sure if this situation could ever has previously
4944 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4947 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4948 header. This is something that is recommended in a new Internet Draft, and
4949 is something that is documented as being done by Sendmail. There are two
4950 possible values. For messages generated by the autoreply transport, Exim
4953 Auto-Submitted: auto-replied
4955 whereas for all other generated messages (e.g. bounces) it adds
4957 Auto-Submitted: auto-generated
4959 63. The "personal" condition in filters now includes a test for the
4960 Auto-Submitted: header. If it contains the string "auto-" the message it
4961 not considered personal.
4963 64. Added rcpt_include_affixes as a generic transport option.
4965 65. Added queue_only_override (default true).
4967 66. Added the syslog_duplication option.
4969 67. If what should have been the first header line of a message consisted of
4970 a space followed by a colon, Exim was mis-interpreting it as a header line.
4971 It isn't of course - it is syntactically invalid and should therefore be
4972 treated as the start of the message body. The misbehaviour could have
4973 caused a number of strange effects, including loss of data in subsequent
4974 header lines, and spool format errors.
4976 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4977 client host had authenticated. This control can now be exercised by an ACL
4978 for more flexibility.
4980 69. By default, callouts do not happen when testing with -bh. There is now a
4981 variant, -bhc, which does actually run the callout code, including
4982 consulting and updating the callout cache.
4984 70. Added support for saslauthd authentication, courtesy of Alexander
4987 71. If statvfs() failed on the spool or log directories while checking their
4988 size for availability, Exim confusingly gave the error "space shortage".
4989 Furthermore, in debugging mode it crashed with a floating point exception.
4990 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4991 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4992 serious problem, Exim now writes to the main and panic logs when this
4993 happens, with details of the failure. It then refuses to accept the
4994 incoming message, giving the message "spool directory problem" or "log
4995 directory problem" with a 421 code for SMTP messages.
4997 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4998 0, 1, and 2 exist, because some OS complain for execs without them (see
4999 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5000 descriptors. However, the code omitted to check that the open succeeded,
5001 causing mysterious errors if for some reason the permissions on /dev/null
5002 got screwed. Now Exim writes a message to the main and panic logs, and
5003 bombs out if it can't open /dev/null.
5005 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5006 interact so that it is all more flexible. It is supposed to remain
5007 backwards compatible. Also added extra_local_interfaces.
5009 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5010 to bomb out with an assertion failure - to the client this appears as a
5011 connection drop. This problem occurs in the part of the code that was taken
5012 from the Samba project. Fortunately, the assertion is in a very simple
5013 function, so I have fixed this by reproducing the function inline in the
5014 one place where it is called, and arranging for authentication to fail
5015 instead of killing the process with assert().
5017 75. The SPA client code was not working when the server requested OEM rather
5018 than Unicode encoding.
5020 76. Added code to make require_files with a specific uid setting more usable in
5021 the case where statting the file as root fails - usually a non-root-mounted
5022 NFS file system. When this happens and the failure is EACCES, Exim now
5023 forks a subprocess and does the per-uid checking as the relevant uid.
5025 77. Added process_log_path.
5027 78. If log_file_path was not explicitly set, a setting of check_log_space or
5028 check_log_inodes was ignored.
5030 79. If a space check for the spool or log partitions fails, the incident is now
5031 logged. Of course, in the latter case the data may get lost...
5033 80. Added the %p formatting code to string_format() so that it can be used to
5034 print addresses in debug_print(). Adjusted all the address printing in the
5035 debugging in store.c to use %p rather than %d.
5037 81. There was a concern that a line of code in smtp_in.c could overflow a
5038 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5039 initially expressed, the concern was not well-founded, because trailing
5040 spaces are removed early. However, if the trailing spaces were followed by
5041 a NULL, they did not get removed, so the overflow was possible. Two fixes
5044 (a) I re-wrote the offending code in a cleaner fashion.
5045 (b) If an incoming SMTP command contains a NULL character, it is rejected
5048 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5049 runs initgroups(), so that if the Exim user is in any additional groups,
5050 they will be used during message reception.
5056 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob