1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.610 2010/05/29 19:16:50 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.
26 JJ/03 installed exipick 20100323.0, fixing doc bug
28 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
29 directory. Notification and patch from Dan Rosenberg
31 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
33 TK/02 Improve log output when DKIM signing operation fails.
35 MH/01 Treat the transport option dkim_domain as a colon separated
36 list, not as a single string, and sign the message with each element,
37 omitting multiple occurences of the same signer.
39 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
40 Bugzilla 985, 986. Patch by Simon Arlott
46 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body
48 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
50 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
52 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
54 NM/04 Bugzilla 915: Flag broken perl installation during build.
60 TK/01 Added patch by Johannes Berg that expands the main option
61 "spamd_address" if it starts with a dollar sign.
63 TK/02 Write list of recipients to X-Envelope-Sender header when building
64 the mbox-format spool file for content scanning (suggested by Jakob
67 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
68 (http://www.dcc-servers.net/) support via dccifd. Activated by
69 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
71 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
72 by Mark Daniel Reidel <mr@df.eu>.
74 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
75 When building exim an external PCRE library is now needed -
76 PCRE is a system library on the majority of modern systems.
77 See entry on PCRE_LIBS in EDITME file.
79 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
80 conversation. Added nologin parameter to request.
81 Patch contributed by Kirill Miazine
83 TF/01 Do not log submission mode rewrites if they do not change the address.
85 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
87 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
88 log files in place. Contributed by Roberto Lima
90 NM/04 Bugzilla 667: close socket used by dovecot authenticator
92 TF/03 Bugzilla 615: When checking the local_parts router precondition
93 after a local_part_suffix or local_part_prefix option, Exim now
94 does not use the address's named list lookup cache, since this
95 contains cached lookups for the whole local part.
97 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
98 Robert Millan. Documentation is in experimental-spec.txt
100 TF/04 Bugzilla 668: Fix parallel build (make -j).
102 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000
104 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
105 Patch provided by Jan Srzednicki
107 TF/05 Leading white space used to be stripped from $spam_report which
108 wrecked the formatting. Now it is preserved.
110 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
111 that they are available at delivery time.
113 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
115 TF/08 TLS error reporting now respects the incoming_interface and
116 incoming_port log selectors.
118 TF/09 Produce a more useful error message if an SMTP transport's hosts
119 setting expands to an empty string.
121 NM/06 Bugzilla 744: EXPN did not work under TLS.
122 Patch provided by Phil Pennock
124 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
125 Patch provided by Richard Godbee
127 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
128 acl_smtp_notquit, added index entry.
130 NM/09 Bugzilla 787: Potential buffer overflow in string_format
131 Patch provided by Eugene Bujak
133 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
134 Patch provided by Maxim Dounin
136 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
137 Patch provided by Phil Pennock
139 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
141 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
142 Patch provided by Brad "anomie" Jorsch
144 NM/14 Bugzilla 622: Added timeout setting to transport filter.
145 Patch provided by Dean Brooks
147 TK/05 Add native DKIM support (does not depend on external libraries).
149 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
150 Patch provided by Graeme Fowler
152 NM/16 Bugzilla 851: Documentation example syntax fix.
154 NM/17 Changed NOTICE file to remove references to embedded PCRE.
156 NM/18 Bugzilla 894: Fix issue with very long lines including comments in lsearch
158 NM/19 Bugzilla 745: TLS version reporting
159 Patch provided by Phil Pennock
161 NM/20 Bugzilla 167: bool: condition support
162 Patch provided by Phil Pennock
164 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken clients
165 Patch provided by Phil Pennock
167 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date
168 Patch provided by Brad "anomie" Jorsch
170 NM/23 Bugzilla 687: Fix misparses in eximstats
171 Patch provided by Heiko Schlittermann
173 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid
174 Patch provided by Heiko Schlittermann
176 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file
177 plus update to original patch.
179 NM/26 Bugzilla 799: Documentation correction for ratelimit
181 NM/27 Bugzilla 802: Improvements to local interface IP addr detection
182 Patch provided by David Brownlee
184 NM/28 Bugzilla 807: Improvements to LMTP delivery logging
186 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes
188 NM/30 Bugzilla 888: TLS documentation bugfixes
190 NM/31 Bugzilla 896: Dovecot buffer overrun fix
192 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
193 Unlike the original bugzilla I have changed all shell scripts in src tree
195 NM/33 Bugzilla 898: Transport filter timeout fix
196 Patch by Todd Rinaldo
198 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches
199 Patch by Serge Demonchaux
201 NM/35 Bugzilla 39: Base64 decode bug fixes
202 Patch by Jakob Hirsch
204 NM/36 Bugzilla 909: Correct connect() call in dcc code
206 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling
208 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed
210 NM/39 Bugzilla 911: Fixed MakeLinks build script
216 TK/01 Add preliminary DKIM support. Currently requires a forked version of
217 ALT-N's libdkim that I have put here:
218 http://duncanthrax.net/exim-experimental/
220 Note to Michael Haardt: I had to rename some vars in sieve.c. They
221 were called 'true' and it seems that C99 defines that as a reserved
222 keyword to be used with 'bool' variable types. That means you could
223 not include C99-style headers which use bools without triggering
224 build errors in sieve.c.
226 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
227 as mailq or other aliases. Changed the --help handling significantly
228 to do whats expected. exim_usage() emits usage/help information.
230 SC/01 Added the -bylocaldomain option to eximstats.
232 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
234 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
236 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
242 PH/01 Another patch from the Sieve maintainer.
244 PH/02 When an IPv6 address is converted to a string for single-key lookup
245 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
246 dots are used instead of colons so that keys in lsearch files need not
247 contain colons. This was done some time before quoting was made available
248 in lsearch files. However, iplsearch files do require colons in IPv6 keys
249 (notated using the quote facility) so as to distinguish them from IPv4
250 keys. This meant that lookups for IP addresses in host lists did not work
251 for iplsearch lookups.
253 This has been fixed by arranging for IPv6 addresses to be expressed with
254 colons if the lookup type is iplsearch. This is not incompatible, because
255 previously such lookups could never work.
257 The situation is now rather anomolous, since one *can* have colons in
258 ordinary lsearch keys. However, making the change in all cases is
259 incompatible and would probably break a number of configurations.
261 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
264 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
265 conversion specification without a maximum field width, thereby enabling
266 a rogue spamd server to cause a buffer overflow. While nobody in their
267 right mind would setup Exim to query an untrusted spamd server, an
268 attacker that gains access to a server running spamd could potentially
269 exploit this vulnerability to run arbitrary code as the Exim user.
271 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
272 $primary_hostname instead of what libspf2 thinks the hosts name is.
274 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
275 a directory entry by the name of the lookup key. Previously, if a
276 symlink pointed to a non-existing file or a file in a directory that
277 Exim lacked permissions to read, a lookup for a key matching that
278 symlink would fail. Now it is enough that a matching directory entry
279 exists, symlink or not. (Bugzilla 503.)
281 PH/03 The body_linecount and body_zerocount variables are now exported in the
284 PH/04 Added the $dnslist_matched variable.
286 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
287 This means they are set thereafter only if the connection becomes
290 PH/06 Added the client_condition to authenticators so that some can be skipped
291 by clients under certain conditions.
293 PH/07 The error message for a badly-placed control=no_multiline_responses left
294 "_responses" off the end of the name.
296 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
298 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
299 (without spaces) instead of just copying the configuration text.
301 PH/10 Added the /noupdate option to the ratelimit ACL condition.
303 PH/11 Added $max_received_linelength.
305 PH/12 Added +ignore_defer and +include_defer to host lists.
307 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
308 way in which PCRE > 7.0 is built.
310 PH/14 Implemented queue_only_load_latch.
312 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
313 MAIL command. The effect was to mangle the value on 64-bit systems.
315 PH/16 Another patch from the Sieve maintainer.
317 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
319 PH/18 If a system quota error occurred while trying to create the file for
320 a maildir delivery, the message "Mailbox is full" was not appended to the
321 bounce if the delivery eventually timed out. Change 4.67/27 below applied
322 only to a quota excession during the actual writing of the file.
324 PH/19 It seems that peer DN values may contain newlines (and other non-printing
325 characters?) which causes problems in log lines. The DN values are now
326 passed through string_printing() before being added to log lines.
328 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
329 and InterBase are left for another time.)
331 PH/21 Added message_body_newlines option.
333 PH/22 Guard against possible overflow in moan_check_errorcopy().
335 PH/23 POSIX allows open() to be a macro; guard against that.
337 PH/24 If the recipient of an error message contained an @ in the local part
338 (suitably quoted, of course), incorrect values were put in $domain and
339 $local_part during the evaluation of errors_copy.
345 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
346 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
349 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
350 issue a MAIL command.
352 PH/02 In an ACL statement such as
354 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
356 if a client was not listed at all, or was listed with a value other than
357 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
358 the condition was not true (as it should be), so access was not denied.
359 The bug was that the ! inversion was incorrectly passed on to the second
360 item. This has been fixed.
362 PH/03 Added additional dnslists conditions == and =& which are different from
363 = and & when the dns lookup returns more than one IP address.
365 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
366 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
368 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
369 FSYNC, which compiles an option called disable_fsync that allows for
370 bypassing fsync(). The documentation is heavily laced with warnings.
372 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
374 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
375 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
376 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
377 including adding "make clean"; (3) Added -fPIC when compiling the test
378 dynamically loaded module, to get rid of a warning.
380 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
381 message fails, move_frozen_messages = true and ignore_bounce_errors_after
382 = 0s. The bug is otherwise harmless.
384 PH/07 There was a bug in the dovecot authenticator such that the value of
385 $auth1 could be overwritten, and so not correctly preserved, after a
386 successful authentication. This usually meant that the value preserved by
387 the server_setid option was incorrect.
389 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
391 PH/09 Installed PCRE release 7.0.
393 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
394 run for batched SMTP input. It is now run at the start of every message
395 in the batch. While fixing this I discovered that the process information
396 (output by running exiwhat) was not always getting set for -bs and -bS
397 input. This is fixed, and it now also says "batched" for BSMTP.
399 PH/11 Added control=no_pipelining.
401 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
402 patch, slightly modified), and move the expansion of helo_data till after
403 the connection is made in the smtp transport (so it can use these
406 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
408 PH/14 Added log_selector = +pid.
410 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
412 PH/16 Add ${if forany and ${if forall.
414 PH/17 Added dsn_from option to vary the From: line in DSNs.
416 PH/18 Flush SMTP output before performing a callout, unless control =
417 no_callout_flush is set.
419 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
420 was true (the default) a successful delivery failed to delete the retry
421 item, thus causing premature timeout of the address. The bug is now
424 PH/20 Added hosts_avoid_pipelining to the smtp transport.
426 PH/21 Long custom messages for fakedefer and fakereject are now split up
427 into multiline reponses in the same way that messages for "deny" and
428 other ACL rejections are.
430 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
431 with slight modification.
433 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
434 draft, changing the syntax and factoring some duplicate code".
436 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
437 for deliveries of the second and subsequent messages over the same SMTP
440 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
441 ${reduce, with only minor "tidies".
443 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
445 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
446 expansion side effects.
448 PH/27 When a message times out after an over-quota error from an Exim-imposed
449 quota, the bounce message says "mailbox is full". This message was not
450 being given when it was a system quota that was exceeded. It now should
453 MH/03 Made $recipients available in local_scan(). local_scan() already has
454 better access to the recipient list through recipients_list[], but
455 $recipients can be useful in postmaster-provided expansion strings.
457 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
458 in the case of a MAIL command with additional options following the
459 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
460 were accidentally chopped off.
462 PH/29 SMTP synchronization checks are implemented when a command is read -
463 there is a check that no more input is waiting when there shouldn't be
464 any. However, for some commands, a delay in an ACL can mean that it is
465 some time before the response is written. In this time, more input might
466 arrive, invalidly. So now there are extra checks after an ACL has run for
467 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
468 pipelining has not been advertised.
470 PH/30 MH's patch to allow iscntrl() characters to be list separators.
472 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
473 returned in the SMTP response when smtp_return_error_details was false.
476 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
477 instead of the C I/O that was originally supplied, because problems were
480 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
481 Exim which did not show up earlier: it was assuming that a call to
482 SSL_CTX_set_info_callback() might give an error value. In fact, there is
483 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
484 was a macro that became an assignment, so it seemed to work. This has
485 changed to a proper function call with a void return, hence the compile
486 error. Exim's code has been fixed.
488 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
491 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
493 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
495 PH/37 If a message is not accepted after it has had an id assigned (e.g.
496 because it turns out to be too big or there is a timeout) there is no
497 "Completed" line in the log. When some messages of this type were
498 selected by exigrep, they were listed as "not completed". Others were
499 picked up by some special patterns. I have improved the selection
500 criteria to be more general.
502 PH/38 The host_find_failed option in the manualroute router can now be set
503 to "ignore", to completely ignore a host whose IP address cannot be
504 found. If all hosts are ignored, the behaviour is controlled by the new
505 host_all_ignored option.
507 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
508 homing or because of multiple MX records with /mx) generated more than
509 one IP address, and the following item turned out to be the local host,
510 all the secondary addresses of the first item were incorrectly removed
511 from the list, along with the local host and any following hosts (which
512 is what is supposed to happen).
514 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
515 whoever called Exim into the -H file. In the case of the daemon it was
516 behaving confusingly. When first started, it used values for whoever
517 started the daemon, but after a SIGHUP it used the Exim user (because it
518 calls itself on a restart). I have changed the code so that it now always
521 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
522 message are rejected with the same error (e.g. no authentication or bad
523 sender address), and a DATA command is nevertheless sent (as can happen
524 with PIPELINING or a stupid MUA), the error message that was given to the
525 RCPT commands is included in the rejection of the DATA command. This is
526 intended to be helpful for MUAs that show only the final error to their
529 PH/42 Another patch from the Sieve maintainer.
531 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
532 Eximstats - Fixed some broken HTML links and added missing column headers
534 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
535 columns for Rejects, Temp Rejects, Ham, and Spam rows.
537 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
539 PH/43 Yet another patch from the Sieve maintainer.
541 PH/44 I found a way to check for a TCP/IP connection going away before sending
542 the response to the final '.' that terminates a message, but only in the
543 case where the client has not sent further data following the '.'
544 (unfortunately, this is allowed). However, in many cases there won't be
545 any further data because there won't be any more messages to send. A call
546 to select() can be used: if it shows that the input is "ready", there is
547 either input waiting, or the socket has been closed. An attempt to read
548 the next input character can distinguish the two cases. Previously, Exim
549 would have sent an OK response which the client would never have see.
550 This could lead to message repetition. This fix should cure that, at
551 least in a lot of common cases.
553 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
554 advertised in response to EHLO.
560 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
561 fixed by 4.65/MH/01 (is this a record?) are fixed:
563 (i) An empty string was always treated as zero by the numeric comparison
564 operators. This behaviour has been restored.
566 (ii) It is documented that the numeric comparison operators always treat
567 their arguments as decimal numbers. This was broken in that numbers
568 starting with 0 were being interpreted as octal.
570 While fixing these problems I realized that there was another issue that
571 hadn't been noticed. Values of message_size_limit (both the global option
572 and the transport option) were treated as octal if they started with 0.
573 The documentation was vague. These values are now always treated as
574 decimal, and I will make that clear in the documentation.
580 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
581 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
584 MH/01 Don't check that the operands of numeric comparison operators are
585 integers when their expansion is in "skipping" mode (fixes bug
586 introduced by 4.64-PH/07).
588 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
589 child addresses, Exim now panics and dies. Previously, because the count
590 is held in a short int, deliveries were likely to be lost. As such a
591 large number of recipients for a single message is ridiculous
592 (performance will be very, very poor), I have chosen to impose a limit
593 rather than extend the field.
599 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
600 leftover -K file (the existence of which was triggered by #402).
601 While we were at it, introduced process PID as part of the -K
602 filename. This should rule out race conditions when creating
605 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
606 processing considerably. Previous code took too long for large mails,
607 triggering a timeout which in turn triggers #401.
609 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
610 in the DK code in transports.c. sendfile() is not really portable,
611 hence the _LINUX specificness.
613 TF/01 In the add_headers option to the mail command in an Exim filter,
614 there was a bug that Exim would claim a syntax error in any
615 header after the first one which had an odd number of characters
618 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
619 callout verification, Exim cached a "reject" for the entire domain. This
620 is correct for most verifications, but it is not correct for a recipient
621 verification with use_sender or use_postmaster set, because in that case
622 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
623 case of MAIL FROM:<> rejection from other early rejections (e.g.
624 rejection of HELO). When verifying a recipient using a non-null MAIL
625 address, the cache is ignored if it shows MAIL FROM:<> rejection.
626 Whatever the result of the callout, the value of the domain cache is
627 left unchanged (for any other kind of callout, getting as far as trying
628 RCPT means that the domain itself is ok).
630 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
633 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
634 manifest itself as EPIPE rather than ECONNECT. When tidying away a
635 session, the daemon ignores ECONNECT errors and logs others; it now
636 ignores EPIPE as well.
638 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
639 (quoted-printable decoding).
641 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
642 later the small subsequent patch to fix an introduced bug.
644 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
646 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
648 PH/08 An error is now given if message_size_limit is specified negative.
650 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
651 to be given (somewhat) arbitrary names.
653 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
656 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
657 miscellaneous code fixes
659 PH/10 Added the log_reject_target ACL modifier to specify where to log
662 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
663 hostname. This is wrong, because it relates to the incoming message (and
664 probably the interface on which it is arriving) and not to the outgoing
665 callout (which could be using a different interface). This has been
666 changed to use the value of the helo_data option from the smtp transport
667 instead - this is what is used when a message is actually being sent. If
668 there is no remote transport (possible with a router that sets up host
669 addresses), $smtp_active_hostname is used.
671 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
672 tweaks were necessary in order to get it to work (see also 21 below):
673 (a) The code assumed that strncpy() returns a negative number on buffer
674 overflow, which isn't the case. Replaced with Exim's string_format()
676 (b) There were several signed/unsigned issues. I just did the minimum
677 hacking in of casts. There is scope for a larger refactoring.
678 (c) The code used strcasecmp() which is not a standard C function.
679 Replaced with Exim's strcmpic() function.
680 (d) The code set only $1; it now sets $auth1 as well.
681 (e) A simple test gave the error "authentication client didn't specify
682 service in request". It would seem that Dovecot has changed its
683 interface. Fortunately there's a specification; I followed it and
684 changed what the client sends and it appears to be working now.
686 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
689 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
690 address A is aliased to B and C, where B exists and C does not. Without
691 -v the output is "A verified" because verification stops after a
692 successful redirection if more than one address is generated. However,
693 with -v the child addresses are also verified. Exim was outputting "A
694 failed to verify" and then showing the successful verification for C,
695 with its parentage. It now outputs "B failed to verify", showing B's
696 parentage before showing the successful verification of C.
698 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
699 look up a TXT record in a specific list after matching in a combined
702 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
703 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
704 they consult the DNS. I had assumed they would set it the way they
705 wanted; and indeed my experiments on Linux seem to show that in some
706 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
707 To be on the safe side, however, I have now made the interface to
708 host_find_byname() similar to host_find_bydns(), with an argument
709 containing the DNS resolver options. The host_find_byname() function now
710 sets these options at its start, just as host_find_bydns() does. The smtp
711 transport options dns_qualify_single and dns_search_parents are passed to
712 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
713 of host_find_byname() use the default settings of RES_DEFNAMES
714 (qualify_single) but not RES_DNSRCH (search_parents).
716 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
717 spool_read_header() do less string testing, by means of a preliminary
718 switch on the second character of optional "-foo" lines. (This is
719 overdue, caused by the large number of possibilities that now exist.
720 Originally there were few.) While I was there, I also converted the
721 str(n)cmp tests so they don't re-test the leading "-" and the first
722 character, in the hope this might squeeze out yet more improvement.
724 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
725 flag allowing group syntax was set by the header_syntax check but not
726 turned off, possible causing trouble later; (2) The flag was not being
727 set at all for the header_verify test, causing "group"-style headers to
728 be rejected. I have now set it in this case, and also caused header_
729 verify to ignore an empty address taken from a group. While doing this, I
730 came across some other cases where the code for allowing group syntax
731 while scanning a header line wasn't quite right (mostly, not resetting
732 the flag correctly in the right place). These bugs could have caused
733 trouble for malformed header lines. I hope it is now all correct.
735 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
736 with the "reply" argument non-NULL. The code, however (which originally
737 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
738 but it didn't always do it. This confused somebody who was copying the
739 code for some other use. I have removed all the tests.
741 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
742 feature that was used to support insecure browsers during the U.S. crypto
743 embargo. It requires special client support, and Exim is probably the
744 only MTA that supported it -- and would never use it because real RSA is
745 always available. This code has been removed, because it had the bad
746 effect of slowing Exim down by computing (never used) parameters for the
747 RSA_EXPORT functionality.
749 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
750 authenticator to fail if there's a tab character in the incoming data
751 (there should never be unless someone is messing about, as it's supposed
752 to be base64-encoded). Also added, on Timo's advice, the "secured" option
753 if the connection is using TLS or if the remote IP is the same as the
754 local IP, and the "valid-client-cert option" if a client certificate has
757 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
758 authenticators. This can be used for authorization after authentication
759 succeeds. (In the case of plaintext, it servers for both authentication
762 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
763 if any retry times were supplied.
765 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
766 connection, where there is no client IP address to check. In this
767 situation, the verify now always succeeds.
769 PH/25 Applied John Jetmore's -Mset patch.
771 PH/26 Added -bem to be like -Mset, but loading a message from a file.
773 PH/27 In a string expansion for a processed (not raw) header when multiple
774 headers of the same name were present, leading whitespace was being
775 removed from all of them, but trailing whitespace was being removed only
776 from the last one. Now trailing whitespace is removed from each header
777 before concatenation. Completely empty headers in a concatenation (as
780 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
781 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
783 PH/29 [Removed. This was a change that I later backed out, and forgot to
784 correct the ChangeLog entry (that I had efficiently created) before
785 committing the later change.]
787 PH/30 Exim was sometimes attempting to deliver messages that had suffered
788 address errors (4xx response to RCPT) over the same connection as other
789 messages routed to the same hosts. Such deliveries are always "forced",
790 so retry times are not inspected. This resulted in far too many retries
791 for the affected addresses. The effect occurred only when there were more
792 hosts than the hosts_max_try setting in the smtp transport when it had
793 the 4xx errors. Those hosts that it had tried were not added to the list
794 of hosts for which the message was waiting, so if all were tried, there
795 was no problem. Two fixes have been applied:
797 (i) If there are any address or message errors in an SMTP delivery, none
798 of the hosts (tried or untried) are now added to the list of hosts
799 for which the message is waiting, so the message should not be a
800 candidate for sending over the same connection that was used for a
801 successful delivery of some other message. This seems entirely
802 reasonable: after all the message is NOT "waiting for some host".
803 This is so "obvious" that I'm not sure why it wasn't done
804 previously. Hope I haven't missed anything, but it can't do any
805 harm, as the worst effect is to miss an optimization.
807 (ii) If, despite (i), such a delivery is accidentally attempted, the
808 routing retry time is respected, so at least it doesn't keep
809 hammering the server.
811 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
812 in ${readsocket because some servers need this prod.
814 PH/32 Added some extra debug output when updating a wait-xxx database.
816 PH/33 The hint "could be header name not terminated by colon", which has been
817 given for certain expansion errors for a long time, was not being given
818 for the ${if def:h_colon_omitted{... case.
820 PH/34 The spec says: "With one important exception, whenever a domain list is
821 being scanned, $domain contains the subject domain." There was at least
822 one case where this was not true.
824 PH/35 The error "getsockname() failed: connection reset by peer" was being
825 written to the panic log as well as the main log, but it isn't really
826 panic-worthy as it just means the connection died rather early on. I have
827 removed the panic log writing for the ECONNRESET error when getsockname()
830 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
831 runs only) independently of the message's sender address. This meant
832 that, if the 4xx error was in fact related to the sender, a different
833 message to the same recipient with a different sender could confuse
834 things. In particualar, this can happen when sending to a greylisting
835 server, but other circumstances could also provoke similar problems.
836 I have changed the default so that the retry time for these errors is now
837 based a combination of the sender and recipient addresses. This change
838 can be overridden by setting address_retry_include_sender=false in the
841 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
842 remote server are returned as part of bounce messages. This was not
843 happening for LMTP over a pipe (the lmtp transport), but now it is the
844 same for both kinds of LMTP.
846 PH/38 Despite being documented as not happening, Exim was rewriting addresses
847 in header lines that were in fact CNAMEs. This is no longer the case.
849 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
850 and queue runs started by the daemon processed all messages. This has
851 been fixed so that -R and -S can now usefully be given with -q<time>.
853 PH/40 Import PCRE release 6.7 (fixes some bugs).
855 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
857 PH/42 Give an error if -q is specified more than once.
859 PH/43 Renamed the variables $interface_address and $interface_port as
860 $received_ip_address and $received_port, to make it clear that these
861 values apply to message reception, and not to the outgoing interface when
862 a message is delivered. (The old names remain recognized, of course.)
864 PH/44 There was no timeout on the connect() call when using a Unix domain
865 socket in the ${readsocket expansion. There now is.
867 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
868 be meaningful with "accept".
870 SC/01 Eximstats V1.43
871 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
873 SC/02 Eximstats V1.44
874 Use a glob alias rather than an array ref in the generated
875 parser. This improves both readability and performance.
877 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
878 Collect SpamAssassin and rejection statistics.
879 Don't display local sender or destination tables unless
880 there is data to show.
881 Added average volumes into the top table text output.
883 SC/04 Eximstats V1.46
884 Collect data on the number of addresses (recipients)
885 as well as the number of messages.
887 SC/05 Eximstats V1.47
888 Added 'Message too big' to the list of mail rejection
889 reasons (thanks to Marco Gaiarin).
891 SC/06 Eximstats V1.48
892 Mainlog lines which have GMT offsets and are too short to
893 have a flag are now skipped.
895 SC/07 Eximstats V1.49 (Alain Williams)
896 Added the -emptyok flag.
898 SC/08 Eximstats V1.50
899 Fixes for obtaining the IP address from reject messages.
901 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
902 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
903 whitesspace changes from 4.64-PH/27
905 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
908 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
909 are found, allow negative numbers in numeric criteria)
911 JJ/06 exipick.20061117.2, added new $message_body_missing variable
913 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
914 to match changes made in 4.64-PH/43
916 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
918 PH/47 Put in an explicit test for a DNS lookup of an address record where the
919 "domain" is actually an IP address, and force a failure. This locks out
920 those revolvers/nameservers that support "A-for-A" lookups, in
921 contravention of the specifications.
923 PH/48 When a host name was looked up from an IP address, and the subsequent
924 forward lookup of the name timed out, the host name was left in
925 $sender_host_name, contrary to the specification.
927 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
928 restricted to single-key lookups, Exim was not diagnosing an error if
929 * or *@ was used with a query-style lookup.
931 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
933 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
934 long ago, but noone interested enough thought of it. Let's just say that
935 the "1.1" means that there are some new functions that weren't there at
936 some point in the past.
938 PH/51 Error processing for expansion failure of helo_data from an smtp
939 transport during callout processing was broken.
941 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
942 tested/used via the -bh/-bhc/-bs options.
944 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
945 bug, fixed in subsequent PCRE releases).
947 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
948 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
950 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
956 SC/01 Use a glob alias rather than an array ref in eximstats generated
957 parser. This improves both readability and performance.
959 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
960 Don't display local sender or destination tables in eximstats unless
961 there is data to show.
962 Added average volumes into the eximstats top table text output.
964 SC/03 Collect data on the number of addresses (recipients) as well
965 as the number of messages in eximstats.
967 TF/01 Correct an error in the documentation for the redirect router. Exim
968 does (usually) call initgroups() when daemonizing.
970 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
971 with consistent privilege compared to when running as a daemon.
973 TF/03 Note in the spec that $authenticated_id is not set for local
974 submissions from trusted users.
976 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
977 Thanks to Dean Brooks <dean@iglou.com> for the patch.
979 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
980 by adding some example configuration directives to the default
981 configuration file. A little bit of work is required to uncomment the
982 directives and define how usernames and passwords are checked, but
983 there is now a framework to start from.
985 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
986 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
987 without this. I don't know how relevant this is to other LDAP libraries.
989 PH/02 Add the verb name to the "unknown ACL verb" error.
991 PH/03 Magnus Holmgren's patch for filter_prepend_home.
993 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
995 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
996 directory not expanded when it should be if an expanded home directory
997 was set for the address (which is overridden by the transport).
999 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1002 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1003 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1004 because it is too late at that time, and has no effect.
1006 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1007 security issue with \' (bugzilla #107). I could not use the
1008 PQescapeStringConn() function, because it needs a PGconn value as one of
1011 PH/08 When testing addresses using -bt, indicate those final addresses that
1012 are duplicates that would not cause an additional delivery. At least one
1013 person was confused, thinking that -bt output corresponded to deliveries.
1014 (Suppressing duplicates isn't a good idea as you lose the information
1015 about possibly different redirections that led to the duplicates.)
1017 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1018 systems where poll() doesn't work, in particular OS X.
1020 PH/10 Added more information to debugging output for retry time not reached.
1022 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1023 operations in malware.c.
1025 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1028 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1029 syslog_duplication set false, log lines that would normally be written
1030 both the the main log and to the reject log were not written to syslog at
1033 PH/14 In the default configuration, change the use of "message" in ACL warn
1034 statements to "add_header".
1036 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1037 not followed by a command (e.g. "seen endif").
1039 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1040 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1043 PH/17 Added extra conditions to the default value of delay_warning_condition
1047 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1048 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1049 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1052 The Auto-Submitted: and various List- headers are standardised, whereas I
1053 don't think Precedence: ever was.
1055 PH/18 Refactored debugging code in route_finduser() to show more information,
1056 in particular, the error code if getpwnam() issues one.
1058 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1059 This is apparently needed in addition to the PH/07 change above to avoid
1060 any possible encoding problems.
1062 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1063 but not after initializing Perl.
1065 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1066 output them only if debugging. By default they are written stderr,
1067 apparently, which is not desirable.
1069 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1072 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1075 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1077 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1078 authenticated or an ident call has been made. Suppress the default
1079 values for $authenticated_id and $authenticated_sender (but permit -oMai
1080 and -oMas) when testing with -bh.
1082 PH/24 Re-jigged the order of the tests in the default configuration so that the
1083 tests for valid domains and recipients precede the DNS black list and CSA
1084 tests, on the grounds that those ones are more expensive.
1086 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1087 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1088 This bug exists in every version of Exim that I still have, right back to
1091 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1092 However, an attempt to turn on case-sensitivity in a regex key by
1093 including (?-i) didn't work because the subject string was already
1094 lowercased, and the effects were non-intuitive. It turns out that a
1095 one-line patch can be used to allow (?-i) to work as expected.
1101 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1102 other effects) broke the use of negated acl sub-conditions.
1104 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1107 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1108 "Deny" causes Exim to reject the incoming connection with a 554 error.
1109 Unfortunately, if there is a major crisis, such as a disk failure,
1110 tcp-wrappers gives "deny", whereas what one would like would be some
1111 kind of temporary error. A kludge has been added to help with this.
1112 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1113 554 error is used if errno is still zero or contains ENOENT (which occurs
1114 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1117 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1119 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1120 errors. Otherwise a message that provokes a temporary error (when other
1121 messages do not) can cause a whole host to time out.
1123 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1124 addresses were routed directly to files or pipes from a redirect router.
1125 File deliveries just didn't batch; pipe deliveries might have suffered
1128 PH/06 A failure to get a lock for a hints database would erroneously always say
1129 "Failed to get write lock", even when it was really a read lock.
1131 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1132 of 0600. This has been changed to use the value of the lockfile_mode
1133 option (which defaults to 0600).
1135 PH/08 Applied small patch from the Sieve maintainer.
1137 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1138 folder from quota calculations, a direct delivery into this folder messed
1139 up the contents of the maildirsize file. This was because the regex was
1140 used only to exclude .Trash (or whatever) when the size of the mailbox
1141 was calculated. There was no check that a delivery was happening into an
1142 excluded directory. This bug has been fixed by ignoring all quota
1143 processing for deliveries into excluded directories.
1145 PH/10 Added the maildirfolder_create_regex option to appendfile.
1151 PH/01 The code for finding all the local interface addresses on a FreeBSD
1152 system running IPv6 was broken. This may well have applied to all BSD
1153 systems, as well as to others that have similar system calls. The broken
1154 code found IPv4 interfaces correctly, but gave incorrect values for the
1155 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1156 that it would not match correctly against @[] and not recognize the IPv6
1159 PH/02 The ipliteral router was not recognizing addresses of the form user@
1160 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1162 PH/03 Added disable_ipv6.
1164 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1165 decoded form, because it is most often used to construct To: headers
1166 lines in autoreplies, and the decoded form may well be syntactically
1167 invalid. However, $reply_address has leading white space removed, and all
1168 newlines turned into spaces so that the autoreply transport does not
1171 PH/05 If group was specified without a user on a router, and no group or user
1172 was specified on a transport, the group from the router was ignored.
1174 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1175 for visible compile-time settings that can be used to change these
1176 numbers, for those that want even more. Backwards compatibility with old
1177 spool files has been maintained. However, going back to a previous Exim
1178 release will lost any variables that are in spool files.
1180 PH/07 Two small changes when running in the test harness: increase delay when
1181 passing a TCP/IP connection to a new process, in case the original
1182 process has to generate a bounce, and remove special handling of
1183 127.0.0.2 (sic), which is no longer necessary.
1185 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1186 be the same on different OS.
1188 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1191 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1192 whether --show-vars was specified or not
1194 JJ/02 exipick: Added support for new ACL variable spool format introduced
1197 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1198 syntactically invalid From: or Reply-to: line, and a filter used this to
1199 generate an autoreply, and therefore failed to obtain an address for the
1200 autoreply, Exim could try to deliver to a non-existent relative file
1201 name, causing unrelated and misleading errors. What now happens is that
1202 it logs this as a hard delivery error, but does not attempt to create a
1205 PH/11 The exinext utility has a -C option for testing purposes, but although
1206 the given file was scanned by exinext itself; it wasn't being passed on
1207 when Exim was called.
1209 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1210 an end-of-file indication when reading a command response.
1212 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1213 compiled. In many other places in Exim, IPv6 addresses are always
1214 recognized, so I have changed this. It also means that IPv4 domain
1215 literals of the form [IPV4:n.n.n.n] are now always recognized.
1217 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1218 used if the router is not running as root, for example, when verifying at
1219 ACL time, or when using -bh. The debugging output from this situation was
1220 non-existent - all you got was a failure to exec. I have made two
1223 (a) Failures to set uid/gid, the current directory, or a process leader
1224 in a subprocess such as that created by queryprogram now generate
1225 suitable debugging ouput when -d is set.
1227 (b) The queryprogram router detects when it is not running as root,
1228 outputs suitable debugging information if -d is set, and then runs
1229 the subprocess without attempting to change uid/gid.
1231 PH/15 Minor change to Makefile for building test_host (undocumented testing
1234 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1235 additional section of a DNS packet that returns MX or SRV records.
1236 Instead, it always explicitly searches for A/AAAA records. This avoids
1237 major problems that occur when a DNS server includes only records of one
1238 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1239 fixed another bug: if SRV records were looked up and the corresponding
1240 address records were *not* found in the additional section, the port
1241 values from the SRV records were lost.
1243 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1244 using the correct key (the original address) when searching the retry
1245 rules in order to find which one to use for generating the retry hint.
1247 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1248 adding the default one. Similarly, if it contains a Reply-To: header, the
1249 errors_reply_to option, if set, is not used.
1251 PH/19 When calculating a retry time, Exim used to measure the "time since
1252 failure" by looking at the "first failed" field in the retry record. Now
1253 it does not use this if it is later than than the arrival time of the
1254 message. Instead it uses the arrival time. This makes for better
1255 behaviour in cases where some deliveries succeed, thus re-setting the
1256 "first failed" field. An example is a quota failure for a huge message
1257 when small messages continue to be delivered. Without this change, the
1258 "time since failure" will always be short, possible causing more frequent
1259 delivery attempts for the huge message than are intended.
1260 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1262 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1263 $1, $2, $3) because the numerical variables can be reset during some
1264 expansion items (e.g. "match"), thereby losing the authentication data.
1266 PH/21 Make -bV show the size of off_t variables so that the test suite can
1267 decide whether to run tests for quotas > 2G.
1269 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1270 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1271 filecount value is greater than 2G or if a quota value is greater than 2G
1272 on a system where the size of off_t is not greater than 4, a panic error
1275 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1276 never match. The debug and -bh output now contains an explicit error
1277 message indicating a malformed IPv4 address or mask.
1279 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1280 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1281 PH/23 above applies.
1283 PH/25 Do not write to syslog when running in the test harness. The only
1284 occasion when this arises is a failure to open the main or panic logs
1285 (for which there is an explicit test).
1287 PH/26 Added the /no_tell option to "control=freeze".
1289 PH/27 If a host name lookup failed very early in a connection, for example, if
1290 the IP address matched host_lookup and the reverse lookup yielded a name
1291 that did not have a forward lookup, an error message of the form "no IP
1292 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1293 could be logged. Now it outputs the IP address instead of "NULL".
1295 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1296 allows the sender and the authenticated sender to be set when
1297 submitting a message from within Exim. Since child_open_exim() is
1298 documented for local_scan(), the new function should be too.
1300 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1301 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1302 results in an empty string is now treated as unset.
1304 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1306 PH/31 Added sender_verify_fail logging option.
1308 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1309 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1310 tidied the source and removed it altogether.
1312 PH/33 When a queue run was abandoned because the load average was too high, a
1313 log line was always written; now it is written only if the queue_run log
1314 selector is set. In addition, the log line for abandonment now contains
1315 information about the queue run such as the pid. This is always present
1316 in "start" and "stop" lines but was omitted from the "abandon" line.
1318 PH/34 Omit spaces between a header name and the colon in the error message that
1319 is given when verify = headers_syntax fails (if there are lots of them,
1320 the message gets confusing).
1322 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1323 names, as there are now some PTR records that contain slashes. This check
1324 is only to protect against broken name servers that fall over on strange
1325 characters, so the fact that it applies to all lookups doesn't matter.
1327 PH/36 Now that the new test suite is complete, we can remove some of the
1328 special code in Exim that was needed for the old test suite. For example,
1329 sorting DNS records because real resolvers return them in an arbitrary
1330 order. The new test suite's fake resolver always returns records in the
1333 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1334 bounces) except when queue_only is set, to avoid logging races between
1335 the different processes.
1337 PH/38 Panic-die if .include specifies a non-absolute path.
1339 PH/39 A tweak to the "H" retry rule from its user.
1341 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1342 a label. They prevented compilation on older perls.
1344 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1345 a warning to be raised on newish perls.
1347 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1348 on queue. Changes to match documented behaviour of showing count of
1349 messages matching specified criteria.
1351 PH/40 Changed the default ident timeout from 30s to 5s.
1353 PH/41 Added support for the use of login_cap features, on those BSD systems
1354 that have them, for controlling the resources used by pipe deliveries.
1356 PH/42 The content-scanning code uses fopen() to create files in which to put
1357 message data. Previously it was not paying any attention to the mode of
1358 the files. Exim runs with umask(0) because the rest of the code creates
1359 files with open(), and sets the required mode explicitly. Thus, these
1360 files were ending up world-writeable. This was not a big issue, because,
1361 being within the spool directory, they were not world-accessible. I have
1362 created a function called modefopen, which takes an additional mode
1363 argument. It sets umask(777), creates the file, chmods it to the required
1364 mode, then resets the umask. All the relevant calls to fopen() in the
1365 content scanning code have been changed to use this function.
1367 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1368 to 24 hours. This avoids potential overflow problems when processing G
1369 and H retry rules. I suspect nobody ever tinkers with this value.
1371 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1373 PH/45 When the plaintext authenticator is running as a client, the server's
1374 challenges are checked to ensure they are valid base64 strings. By
1375 default, the authentication attempt is cancelled if an invalid string is
1376 received. Setting client_ignore_invalid_base64 true ignores these errors.
1377 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1378 they are received. Thus, the responses can be made to depend on the
1379 challenges. If an invalid string is ignored, an empty string is placed in
1382 PH/46 Messages that are created by the autoreply transport now contains a
1383 References: header, in accordance with RFCs 2822 and 3834.
1385 PH/47 Added authenticated_sender_force to the smtp transport.
1387 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1389 PH/49 Installed latest patch from the Sieve maintainer.
1391 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1392 was also set, the appendfile transport was unnecessarily scanning a
1393 directory of message files (e.g. for maildir delivery) to find the count
1394 of files (along with the size), even though it did not need this
1395 information. It now does the scan only if it needs to find either the
1396 size of the count of files.
1398 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1400 PH/52 Two bugs concerned with error handling when the smtp transport is
1403 (i) Exim was not creating retry information for temporary errors given
1404 for individual recipients after the DATA command when the smtp transport
1405 was used in LMTP mode. This meant that they could be retried too
1406 frequently, and not timed out correctly.
1408 (ii) Exim was setting the flag that allows error details to be returned
1409 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1410 recipients that were returned after the DATA command.
1412 PH/53 This is related to PH/52, but is more general: for any failing address,
1413 when detailed error information was permitted to be returned to the
1414 sender, but the error was temporary, then after the final timeout, only
1415 "retry timeout exceeded" was returned. Now it returns the full error as
1416 well as "retry timeout exceeded".
1418 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1419 do this, and (what is worse) MTAs that accept it.
1421 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1422 will now be deprecated.
1424 PH/56 New os.c-cygwin from the Cygwin maintainer.
1426 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1427 formats (previously only available in exim formats via -bpr, -bpru,
1428 and -bpra. Now also available in native and exiqgrep formats)
1430 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1431 with very large, slow to parse queues
1433 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1435 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1437 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1438 responses to authentication challenges, though it was showing the
1439 challenges; (ii) I've removed the CR characters from the debug output for
1442 PH/58 Allow for the insertion of a newline as well as a space when a string
1443 is turned into more than one encoded-word during RFC 2047 encoding. The
1444 Sieve code now uses this.
1446 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1447 data_4xx, lost_connection, tls_required.
1449 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1450 message was being sent as an SMTP response.
1452 PH/61 Add -l and -k options to exicyclog.
1454 PH/62 When verifying, if an address was redirected to one new address, so that
1455 verification continued, and the new address failed or deferred after
1456 having set something in $address_data, the value of $address_data was not
1457 passed back to the ACL. This was different to the case when no
1458 redirection occurred. The value is now passed back in both cases.
1460 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1461 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1462 use login_cap.h, so on its own it isn't the distinguishing feature. The
1463 new name refers directly to the setclassresources() function.
1465 PH/65 Added configuration files for NetBSD3.
1467 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1469 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1470 is preferred over IPv4.
1472 PH/68 The bounce_return_message and bounce_return_body options were not being
1473 honoured for bounces generated during the reception of non-SMTP messages.
1474 In particular, this applied to messages rejected by the ACL. This bug has
1475 been fixed. However, if bounce_return_message is true and bounce_return_
1476 body is false, the headers that are returned for a non-SMTP message
1477 include only those that have been read before the error was detected.
1478 (In the case of an ACL rejection, they have all been read.)
1480 PH/69 The HTML version of the specification is now built in a directory called
1481 spec_html instead of spec.html, because the latter looks like a path with
1482 a MIME-type, and this confuses some software.
1484 PH/70 Catch two compiler warnings in sieve.c.
1486 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1487 function verify_get_ident() calls ip_connect() to connect a socket, but
1488 if the "connect()" function timed out, ip_connect() used to close the
1489 socket. However, verify_get_ident() also closes the socket later, and in
1490 between Exim writes to the log, which may get opened at this point. When
1491 the socket was closed in ip_connect(), the log could get the same file
1492 descriptor number as the socket. This naturally causes chaos. The fix is
1493 not to close the socket in ip_connect(); the socket should be closed by
1494 the function that creates it. There was only one place in the code where
1495 this was missing, in the iplookup router, which I don't think anybody now
1496 uses, but I've fixed it anyway.
1498 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1499 well as to direct DNS lookups. Otherwise the handling of names in host
1500 lists is inconsistent and therefore confusing.
1506 PH/01 Two changes to the default runtime configuration:
1508 (1) Move the checks for relay_from_hosts and authenticated clients from
1509 after to before the (commented out) DNS black list checks.
1511 (2) Add control=submission to the relay_from_hosts and authenticated
1512 clients checks, on the grounds that messages accepted by these
1513 statements are most likely to be submissions.
1515 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1517 (1) Generate an error if the third argument for the ${prvs expansion is
1520 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1523 (3) Reset the variables that are obtained from the first argument of
1524 ${prvscheck and used in the second argument before leaving the code,
1525 because their memory is reclaimed, so using them afterwards may do
1528 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1529 one (it's much easier than Tom thought :-).
1531 (5) Because of (4), we can now allow for the use of $prvscheck_result
1532 inside the third argument.
1534 PH/03 For some reason, the default setting of PATH when running a command from
1535 a pipe transport was just "/usr/bin". I have changed it to
1538 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1539 anything to be listed in the output from -bV.
1541 PH/05 When a filter generated an autoreply, the entire To: header line was
1542 quoted in the delivery log line, like this:
1544 => >A.N.Other <ano@some.domain> <original@ddress> ...
1546 This has been changed so that it extracts the operative address. There
1547 may be more than one such address. If so, they are comma-separated, like
1550 => >ano@some.domain,ona@other.domain <original@ddress> ...
1552 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1553 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1554 not being looked up in the rDNS to get a host name, Exim was showing the
1555 IP address twice in Received: lines, even though the IP addresses were
1556 identical. For example:
1558 Received: from [1.2.3.4] (helo=[1.2.3.4])
1560 However, if the real host name was known, it was omitting the HELO data
1561 if it matched the actual IP address. This has been tidied up so that it
1562 doesn't show the same IP address twice.
1564 PH/07 When both +timestamp and +memory debugging was on, the value given by
1565 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1566 called by the debug printing, thereby overwriting the timestamp buffer.
1567 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1569 PH/08 When the original message was included in an autoreply transport, it
1570 always said "this is a copy of the message, including all the headers",
1571 even if body_only or headers_only was set. It now gives an appropriate
1574 PH/09 Applied a patch from the Sieve maintainer which:
1576 o fixes some comments
1577 o adds the (disabled) notify extension core
1578 o adds some debug output for the result of if/elsif tests
1579 o points to the current vacation draft in the documentation
1580 and documents the missing references header update
1584 o fixes a bug in processing the envelope test (when testing
1585 multiple envelope elements, the last element determinted the
1588 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1589 Electronic Mail") by including:
1591 Auto-submitted: auto-generated
1593 in the messages that it generates (bounce messages and others, such as
1594 warnings). In the case of bounce messages for non-SMTP mesages, there was
1595 also a typo: it was using "Auto_submitted" (underscore instead of
1596 hyphen). Since every message generated by Exim is necessarily in response
1597 to another message, thes have all been changed to:
1599 Auto-Submitted: auto-replied
1601 in accordance with these statements in the RFC:
1603 The auto-replied keyword:
1605 - SHOULD be used on messages sent in direct response to another
1606 message by an automatic process,
1608 - MUST NOT be used on manually-generated messages,
1610 - MAY be used on Delivery Status Notifications (DSNs) and Message
1611 Disposition Notifications (MDNs),
1613 - MUST NOT be used on messages generated by automatic or periodic
1614 processes, except for messages which are automatic responses to
1617 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1618 to the default Received: header definition.
1620 PH/12 Added log selector acl_warn_skipped (default on).
1622 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1623 variables because (a) they are in the wrong storage pool and (b) even if
1624 they were copied, it wouldn't work properly because of the caching.
1626 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1627 checking when decoding. Apparently there are clients that generate
1628 overlong encoded strings. Why am I not surprised?
1630 PH/15 If the first argument of "${if match_address" was not empty, but did not
1631 contain an "@" character, Exim crashed. Now it writes a panic log message
1632 and treats the condition as false.
1634 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1636 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1637 extension "envelope-auth". The code is finished and in agreement with
1638 other implementations, but there is no documentation so far and in fact,
1639 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1640 not changing the active code.
1642 Print executed "if" and "elsif" statements when debugging is used. This
1643 helps a great deal to understand what a filter does.
1645 Document more things not specified clearly in RFC3028. I had all this
1646 sorted out, when out of a sudden new issues came to my mind. Oops."
1648 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1651 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1652 canonical form (as documented). However, after a host name lookup from
1653 the IP address, check_host() was doing a simple string comparison with
1654 addresses acquired from the DNS when checking that the found name did
1655 have the original IP as one of its addresses. Since any found IPv6
1656 addresses are likely to be in abbreviated form, the comparison could
1657 fail. Luckily, there already exists a function for doing the comparison
1658 by converting both addresses to binary, so now that is used instead of
1659 the text comparison.
1661 PH/20 There was another similar case to PH/19, when a complete host name was
1662 given in a host list; looking up its IP address could give an abbreviated
1663 form, whereas the current host's name might or might not be abbreviated.
1664 The same fix has been applied.
1670 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1671 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1674 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1675 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1677 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1679 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1680 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1681 Also, for "Auto-Submitted", treat anything other than "no" as
1682 non-personal, in accordance with RFC 3834. (Previously it treated
1683 anything starting "auto-" as non-personal.)
1685 TF/01 The control=submission/name=... option had a problem with syntax
1686 errors if the name included a slash character. The /name= option
1687 now slurps the rest of the string, so it can include any characters
1688 but it must come last in the list of options (after /sender_retain
1691 PH/05 Some modifications to the interface to the fake nameserver for the new
1699 TK/01 Added the "success_on_redirect" address verification option. See
1700 NewStuff for rationale and an example.
1702 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1704 PH/02 Patch to exigrep to allow it to work on syslog lines.
1706 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1707 fread() to skip over the body file's header line, because in Cygwin the
1708 header line is locked and is inaccessible.
1710 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1711 co-exist for some time) to make it clear that it is the Exim ID that is
1712 referenced, not the Message-ID: header line.
1714 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1715 string_format() function, because snprintf() does not exist on all
1718 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1719 ${stat: expansion item.
1721 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1722 protocol synchronization error", to keep the pedants happy.
1724 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1725 well as for IRIX systems, when gcc is being used. See the host.c source
1728 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1730 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1733 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1734 transport and to the smtp transport in LMTP mode.
1736 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1738 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1739 run a filter in a subprocess. This could lead to confusion in subsequent
1740 lookups in the parent process. There should also be a search_tidyup() at
1741 the end of the subprocess.
1743 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1744 only if the host matched helo_try_verify_hosts, which caused the
1745 verification to occur when the EHLO/HELO command was issued. The ACL just
1746 tested the remembered result. Now, if a previous verification attempt has
1747 not happened, "verify = helo" does it there and then.
1749 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1751 TK/03 Fix log output including CR from clamd.
1753 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1754 exist provoked a memory error which could cause a segfault.
1756 PH/15 Installed PCRE 6.2
1758 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1760 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1761 of the problem. Specifically, suggested +O2 rather than +O1 for the
1764 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1766 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1767 record, the port was not correctly logged when the outgoing_port log
1768 selector was set (it logged the transort's default port).
1770 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1771 fallback_hosts, and "hosts" in the smtp transport.
1773 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1774 host errors such as "Connection refused".
1776 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1777 authentication with radiusclient 0.4.9:
1779 - Error returned from rc_read_config was caught wrongly
1780 - Username/password not passed on to radius server due to wrong length.
1782 The presumption is that some radiusclient API changes for 4.51/PH/17
1783 were not taken care of correctly. The code is still untested by me (my
1784 Linux distribution still has 0.3.2 of radiusclient), but it was
1785 contributed by a Radius user.
1787 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1788 expanding the "port" option of the smtp transport.
1790 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1791 while reading a MIME header. Thanks to Tom Hughes for a patch.
1793 PH/24 Include config.h inside local_scan.h so that configuration settings are
1796 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1797 that in an ACL for RCPT (for example), you can examine exactly what was
1800 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1801 commands, but it was not correctly comparing the address with the actual
1802 client host address. Thus, it would show the EHLO address in Received:
1803 header lines when this was not necessary.
1805 PH/27 Added the % operator to ${eval:}.
1807 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1808 it should be ignoring failures (because with -C, for example, it has lost
1809 privilege). It wasn't ignoring creation failures other than "already
1812 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1815 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1816 because an input line was too long, either on its own, or by virtue of
1817 too many continuations, the temporary file was not being removed, and the
1818 return code was incorrect.
1820 PH/31 Missing "BOOL" in function definition in filtertest.c.
1822 PH/32 Applied Sieve patches from the maintainer.
1824 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1826 PH/33 Added "verify = not_blind".
1828 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1829 Local/Makefile (with some defaults set). These are used in built scripts
1830 such as exicyclog, but they have never been used in the exim_install
1831 script (though there are many overriding facilities there). I have
1832 arranged that the exim_install script now takes note of these two
1835 PH/35 Installed configuration files for Dragonfly.
1837 PH/36 When a locally submitted message by a trusted user did not contain a
1838 From: header, and the sender address was obtained from -f or from an SMTP
1839 MAIL command, and the trusted user did not use -F to supply a sender
1840 name, $originator_name was incorrectly used when constructing a From:
1841 header. Furthermore, $originator_name was used for submission mode
1842 messages from external hosts without From: headers in a similar way,
1843 which is clearly wrong.
1845 PH/37 Added control=suppress_local_fixups.
1847 PH/38 When log_selector = +received_sender was set, and the addition of the
1848 sender made the log line's construction buffer exactly full, or one byte
1849 less than full, an overflow happened when the terminating "\n" was
1852 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1853 when the result of a list match is failure because a DNS lookup failed.
1855 PH/40 RM_COMMAND is now used in the building process.
1857 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1858 the "build-* directories that it finds.
1860 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1861 address in a domain literal was a prefix of an interface address.
1863 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1864 when verifying a sender address, unless rewrite_headers is false.
1866 PH/44 Wrote a long comment about why errors_to addresses are verified as
1867 recipients, not senders.
1869 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1870 the ratelimit ACL was added.
1872 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1874 PH/46 Added extra information about PostgreSQL errors to the error string.
1876 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1877 avoiding the need to install special zones in a real server. This is
1878 backwards compatible; if it can't find the fake resolver, it drops back.
1879 Thus, both old and new test suites can be run.
1881 TF/02 Added util/ratelimit.pl
1883 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1884 clock is set back in time.
1886 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1887 Candler <B.Candler@pobox.com>.
1889 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1890 OK for addresses that are the result of redirections.
1892 PH/48 A number of further additions for the benefit of the new test suite,
1893 including a fake gethostbyname() that interfaces to the fake DNS resolver
1896 TF/06 The fix for widen_domains has also been applied to qualify_single and
1897 search_parents which are the other dnslookup options that can cause
1900 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1903 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1905 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1906 service" if it encountered an MX record with an empty target hostname.
1907 The message is now "an MX or SRV record indicated no SMTP service".
1909 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1910 if the DNS of the sending site is misconfigured. This is quite a
1911 common situation. This change restores the behaviour of treating a
1912 helo verification defer as a failure.
1914 PH/51 If self=fail was set on a router, the bounce message did not include the
1915 actual error message.
1921 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1923 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1924 command itself ended in error, the underlying message about the transport
1925 filter timeout was being overwritten with the pipe command error. Now the
1926 underlying error message should be appended to the second error message.
1928 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1930 PH/02 Reduce the amount of output that "make" produces by default. Full output
1931 can still be requested.
1933 PH/03 The warning log line about a condition test deferring for a "warn" verb
1934 was being output only once per connection, rather than after each
1935 occurrence (because it was using the same function as for successful
1936 "warn" verbs). This seems wrong, so I have changed it.
1938 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1939 it should not have, which might have caused a crash in the right
1940 circumstances, but probably never did.
1942 PH/04 Installed a modified version of Tony Finch's patch to make submission
1943 mode fix the return path as well as the Sender: header line, and to
1944 add a /name= option so that you can make the user's friendly name appear
1947 TF/03 Added the control = fakedefer ACL modifier.
1949 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1950 Mark Lowes for thorough testing.
1952 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1954 TK/03 Merged latest SRS patch from Miles Wilton.
1956 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1957 with the definition in sysexits.h (which is #included earlier).
1958 Fortunately, Exim does not actually use EX_OK. The code used to try to
1959 preserve the sysexits.h value, by assumimg that macro definitions were
1960 scanned for macro replacements. I have been disabused of this notion,
1961 so now the code just undefines EX_OK before #including unistd.h.
1963 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1964 in the smtp transport. When a block could not be written in a single
1965 write() function, the timeout was being re-applied to each part-write.
1966 This seems wrong - if the receiver was accepting one byte at a time it
1967 would take for ever. The timeout is now adjusted when this happens. It
1968 doesn't have to be particularly precise.
1970 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1971 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1973 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1974 without a domain if the check to <postmaster@domain> fails.
1976 SC/01 Eximstats: added -xls and the ability to specify output files
1977 (patch written by Frank Heydlauf).
1979 SC/02 Eximstats: use FileHandles for outputing results.
1981 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1983 SC/04 Eximstats: fixed display of large numbers with -nvr option
1985 SC/05 Eximstats: fixed merging of reports with empty tables.
1987 SC/06 Eximstats: added the -include_original_destination flag
1989 SC/07 Eximstats: removed tabs and trailing whitespace.
1991 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1993 TK/06 MBOX spool code: Add real "From " MBOX separator line
1994 so the .eml file is really in mbox format (even though
1995 most programs do not really care). Patch from Alex Miller.
1997 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1998 The latter is generated from $received_to and is only set if the
1999 message has one envelope recipient. SA can use these headers,
2000 obviously out-of-the-box. Patch from Alex Miller.
2002 PH/08 The ${def test on a variable was returning false if the variable's
2003 value was "0", contrary to what the specification has always said!
2004 The result should be true unless the variable is empty.
2006 PH/09 The syntax error of a character other than { following "${if
2007 def:variable_name" (after optional whitespace) was not being diagnosed.
2008 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2009 accidental colon was present, for example, could give incorrect results.
2011 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2012 result is used (not including appendfile, where other changes are about
2015 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2016 This involved changing a lot of size variables from int to off_t. It
2017 should work with maildirs and everything.
2019 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2020 spamd dying while we are connected to it.
2022 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2025 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2026 function for BDB 4.3.
2028 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2030 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2031 the sources for PCRE have been re-arranged and more files are now
2034 PH/15 The code I had for printing potentially long long variables in PH/11
2035 above was not the best (it lost precision). The length of off_t variables
2036 is now inspected at build time, and an appropriate printing format (%ld
2037 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2038 to be "long long int" or "long int". This is needed for the internal
2039 formatting function string_vformat().
2041 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2042 the configuration file to be ":syslog", then the script "guesses" where
2043 the logs files are, rather than using the compiled in default. In our
2044 case the guess is not the same as the compiled default, so the script
2045 suddenly stopped working when I started to use syslog. The patch checks
2046 to see if log_file_path is "". If so, it attempts to read it from exim
2047 with no configuration file to get the compiled in version, before it
2048 falls back to the previous guessing code."
2050 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2051 implementing BATV in an Exim configuration. See NewStuff for the gory
2054 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2055 Makefile that are specific to HP-UX.
2057 PH/18 If the "use_postmaster" option was set for a recipient callout together
2058 with the "random" option, the postmaster address was used as the MAIL
2059 FROM address for the random test, but not for the subsequent recipient
2060 test. It is now used for both.
2062 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2063 patch removes a few documentation additions to RFC 3028, because the
2064 latest draft now contains them. It adds the new en;ascii-case comparator
2065 and a new error check for 8bit text in MIME parts. Comparator and
2066 require names are now matched exactly. I enabled the subaddress
2067 extension, but it is not well tested yet (read: it works for me)."
2069 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2070 rework some of the code of TK/09 above to avoid the hardwired use of
2071 "%lld" and "long long". Replaced the call to snprintf() with a call to
2074 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2075 records point to non-existent hosts", "retry timeout exceeded", and
2076 "retry time not reached for any host after a long failure period".
2078 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2079 experimental DomainKeys support:
2081 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2082 (2) On an error such as an illegally used "control", the wrong name for
2083 the control was given.
2085 These problems did NOT occur unless DomainKeys support was compiled.
2087 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2089 PH/24 Added ${if match_ip condition.
2091 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2092 databases so that it will be absolutely obvious if a crash occurs in the
2093 DB library. This is a regular occurrence (often caused by mis-matched
2096 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2097 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2098 picked up on a user's system that detects such things. There doesn't seem
2099 to be a gcc warning option for this - only an attribute that has to be
2100 put on the function's prototype. It seems that in Fedora Core 4 they have
2101 set this on a number of new functions. No doubt there will be more in due
2104 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2105 specify a transport. However, if an address that was verified by such a
2106 router was the subject of a callout, Exim crashed because it tried to
2107 read the rcpt_include_affixes from the non-existent transport. Now it
2108 just assumes that the setting of that option is false. This bug was
2109 introduced by 4.51/PH/31.
2111 PH/28 Changed -d+all to exclude +memory, because that information is very
2112 rarely of interest, but it makes the output a lot bigger. People tend to
2113 do -d+all out of habit.
2115 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2116 code in os-type was giving problems when libc.so lives in lib64, like on
2119 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2120 aren't the modern standard, and it seems that some systems' include files
2121 don't always have them. Exim was already checking for some of the newer
2122 ones like T_AAAA, and defining it itself. I've added checks for all the
2123 record types that Exim uses.
2125 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2126 not automatically generating a new one, as it is supposed to. This
2127 prevented TLS from working. If the file did exist, but contained invalid
2128 data, a new version was generated, as expected. It was only the case of a
2129 non-existent file that was broken.
2131 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2132 with a change in libdomainkeys > 0.64.
2134 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2135 from DNS. If the selector record carries the flag, it now has
2136 precedence over the domain-wide flag.
2138 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2140 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2141 the use of a port name that isn't defined in /etc/services) occurred, the
2142 message was deferred as in a normal delivery, and thus remained on the
2143 spool, instead of being failed because of the mua_wrapper setting. This
2144 is now fixed, and I tidied up some of the mua_wrapper messages at the
2147 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2148 the messages in a hash of arrays rather than using individual hashes.
2149 This is a bit cleaner and results in dramatic memory savings, albeit
2150 at a slight CPU cost.
2152 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2153 as requested by Marc Sherman.
2155 SC/10 Eximstats: added histograms for user specified patterns as requested
2158 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2160 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2161 fopen() in the content-scanning modules that did not already have it.
2167 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2168 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2170 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2172 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2174 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2175 to test Sieve filters that use "vacation".
2177 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2178 that changes the way the GnuTLS parameters are stored in the cache file.
2179 The new format can be generated externally. For backward compatibility,
2180 if the data in the cache doesn't make sense, Exim assumes it has read an
2181 old-format file, and it generates new data and writes a new file. This
2182 means that you can't go back to an older release without removing the
2185 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2186 work if there are any delivery delays because "one_time" forces the
2187 parent to be marked "delivered", so its unseen clone is never tried
2188 again. For this reason, Exim now forbids the simultaneous setting of
2191 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2192 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2193 entry if you want to know the details. The fix, however, neglected to
2194 consider the case when local delivery batching is involved. The test for
2195 "previously delivered" was not happening when checking to see if an
2196 address could be batched with a previous (undelivered) one; under
2197 certain circumstances this could lead to multiple deliveries to the same
2200 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2201 in its include files, and this causes problems building Exim.
2203 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2204 header_syntax) but Exim was just ignoring anything given after a slash.
2205 In particular, this caused confusion with an attempt to use "verify =
2206 reverse_host_lookup/defer_ok". An error is now given when options are
2207 supplied for verify items that do not have them. (Maybe reverse_host_
2208 lookup should have a defer_ok option, but that's a different point.)
2210 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2211 defined by RFC 821) to 2048, because there were problems with some AUTH
2212 commands, and RFC 1869 says the size should be increased for extended
2213 SMTP commands that take arguments.
2215 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2218 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2219 "unknown" error; now it says that the functionality isn't in the binary.
2221 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2222 an address' error message when a string expansion fails (syntax or
2223 whatever). Otherwise the password may appear in the log. Following change
2224 PH/42 below, there is no longer a chance of it appearing in a bounce
2227 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2229 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2230 of its addresses was ever tried. (Bugzilla bug #2.)
2232 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2233 the result incorrectly in the debug output. (It correctly added a newline
2234 to what was transported.)
2236 TF/01 Added $received_time.
2238 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2239 commented out examples of how to interface to a virus scanner and to
2240 SpamAssassin. Also added commented examples of av_scanner and
2241 spamd_address settings.
2243 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2244 and controls are allowed in which ACLs. There were a couple of minor
2245 errors. Some of the entries in the conditions table (which is a table of
2246 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2247 negation of where the condition IS allowed.
2249 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2251 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2252 header file does not have a version number, so I've had to invent a new
2253 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2254 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2255 radiusclient), but it was contributed by a Radius user.
2257 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2258 files or directories, for OpenSSL.
2260 PH/19 When an Exim process that is running as root has to create an Exim log
2261 file, it does so in a subprocess that runs as exim:exim so as to get the
2262 ownership right at creation (otherwise, other Exim processes might see
2263 the file with the wrong ownership). There was no test for failure of this
2264 fork() call, which would lead to the process getting stuck as it waited
2265 for a non-existent subprocess. Forks do occasionally fail when resources
2266 run out. I reviewed all the other calls to fork(); they all seem to check
2269 PH/20 When checking for unexpected SMTP input at connect time (before writing
2270 the banner), Exim was not dealing correctly with a non-positive return
2271 from the read() function. If the client had disconnected by this time,
2272 the result was a log entry for a synchronization error with an empty
2273 string after "input=" when read() returned zero. If read() returned -1
2274 (an event I could not check), uninitialized data bytes were printed.
2275 There were reports of junk text (parts of files, etc) appearing after
2278 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2280 PH/22 Added support for macro redefinition, and (re)definition in between
2281 driver and ACL definitions.
2283 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2284 forgetting to use the resulting value; it was using the unexpanded value.
2286 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2287 hadn't been configured. The fix is from Juergen Kreileder, who
2288 understands it better than I do:
2290 "Here's what I see happening with three configured cyrus_sasl
2291 authenticators configured (plain, login, cram-md5):
2293 On startup auth_cyrus_sasl_init() gets called for each of these.
2294 This means three calls to sasl_listmech() without a specified mech_list.
2295 => SASL tests which mechs of all available mechs actually work
2296 => three warnings about OTP not working
2297 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2299 With the patch, sasl_listmech() also gets called three times. But now
2300 SASL's mech_list option is set to the server_mech specified in the the
2301 authenticator. Or in other words, the answer from sasl_listmech()
2302 gets limited to just the mech you're testing for (which is different
2304 => the return list contains just 'plain' or 'login', 'cram-md5' or
2305 nothing depending on the value of ob->server_mech.
2307 I've just tested the patch: Authentication still works fine,
2308 unavailable mechs specified in the exim configuration are still
2309 caught, and the auth.log warnings about OTP are gone."
2311 PH/25 When debugging is enabled, the contents of the command line are added
2312 to the debugging output, even when log_selector=+arguments is not
2315 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2316 answer is "GNU", and only if the return is "GNU/something" is the answer
2319 PH/27 $acl_verify_message is now set immediately after the failure of a
2320 verification in an ACL, and so is available in subsequent modifiers. In
2321 particular, the message can be preserved by coding like this:
2323 warn !verify = sender
2324 set acl_m0 = $acl_verify_message
2326 Previously, $acl_verify_message was set only while expanding "message"
2327 and "log_message" when a very denied access.
2329 PH/28 Modified OS/os.c-Linux with
2331 -#ifndef OS_LOAD_AVERAGE
2332 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2334 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2335 nomenclature these days.)
2337 PH/29 Installed patch from the Sieve maintainer that adds the options
2338 sieve_useraddress and sieve_subaddress to the redirect router.
2340 PH/30 In these circumstances:
2341 . Two addresses routed to the same list of hosts;
2342 . First host does not offer TLS;
2343 . First host accepts first address;
2344 . First host gives temporary error to second address;
2345 . Second host offers TLS and a TLS session is established;
2346 . Second host accepts second address.
2347 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2348 and peerdn, if requested) that were in fact used only for the second
2351 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2352 attention to any local part prefix or suffix that was matched by the
2353 router that accepted the address. It now behaves in the same way as it
2354 does for delivery: the affixes are removed from the local part unless
2355 rcpt_include_affixes is set on the transport.
2357 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2358 timeout during the DATA phase of an incoming message.
2360 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2361 applied a patch sanctioned by the Sieve maintainer.
2363 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2364 the uid or gid is negative. A case of a negative gid caused this to be
2365 noticed. The fix allows for either to be negative.
2367 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2368 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2371 PH/36 The change PH/12 above was broken. Fixed it.
2373 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2374 the grounds that routing the same address twice would always produce the
2375 same answer. This might have been true once, but it is certainly no
2376 longer true now. Routing a child address may depend on the previous
2377 routing that produced that child. Some complicated redirection strategies
2378 went wrong when messages had multiple recipients, and made Exim's
2379 behaviour dependent on the order in which the addresses were given.
2381 I have moved the duplicate checking until after the routing is complete.
2382 Exim scans the addresses that are assigned to local and remote
2383 transports, and removes any duplicates. This means that more work will be
2384 done, as duplicates will always all be routed, but duplicates are
2385 presumably rare, so I don't expect this is of any significance.
2387 For deliveries to pipes, files, and autoreplies, the duplicate checking
2388 still happens during the routing process, since they are not going to be
2391 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2392 It corrects a timeout issue with spamd. This is Ian's comment: "The
2393 background is that sometimes spamd either never reads data from a
2394 connection it has accepted, or it never writes response data. The exiscan
2395 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2396 blindly assumes that writes won't block so it may never time out."
2398 PH/39 Allow G after quota size as well as K and M.
2400 PH/40 The value set for $authenticated_id in an authenticator may not contain
2401 binary zeroes or newlines because the value is written to log lines and
2402 to spool files. There was no check on this. Now the value is run through
2403 the string_printing() function so that such characters are converted to
2404 printable escape sequences.
2406 PH/41 $message_linecount is a new variable that contains the total number of
2407 lines in the message. Compare $body_linecount, which is the count for the
2410 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2411 bounce and delay warning messages, except in certain special cases, which
2414 (a) An SMTP error message from a remote host;
2415 (b) A message specified in a :fail: redirection;
2416 (c) A message specified in a "fail" command in a system filter;
2417 (d) A message specified in a FAIL return from the queryprogram router;
2418 (e) A message specified by the cannot_route_message router option.
2420 In these cases only, Exim does include the error details in bounce and
2421 warning messages. There are also a few cases where bland messages such
2422 as "unrouteable address" or "local delivery error" are given.
2424 PH/43 $value is now also set for the "else" part of a ${run expansion.
2426 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2427 being worked on, but at least Exim now implements the latest version to
2430 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2431 process to complete was treated as a delivery failure, a timeout while
2432 writing the message to the pipe was logged, but erroneously treated as a
2433 successful delivery. Such timeouts include transport filter timeouts. For
2434 consistency with the overall process timeout, these timeouts are now
2435 treated as errors, giving rise to delivery failures by default. However,
2436 there is now a new Boolean option for the pipe transport called
2437 timeout_defer, which, if set TRUE, converts the failures into defers for
2438 both kinds of timeout. A transport filter timeout is now identified in
2441 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2442 systems where "make" and "gmake" are different, calling "gmake" at top
2443 level broke things. I've arranged for the value of $(MAKE) to be passed
2444 from the Makefile to this script so that it can call the same version of
2448 A note about Exim versions 4.44 and 4.50
2449 ----------------------------------------
2451 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2452 changes of various kinds. As a consequence, a big documentation update was
2453 needed. This delayed the release for rather longer than seemed good, especially
2454 in the light of a couple of (minor) security issues. Therefore, the changes
2455 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2456 release. So 4.44 and 4.50 are in effect two different branches that both start
2459 I have left the 4.50 change log unchanged; it contains all the changes since
2460 4.43. The change log for 4.44 is below; many of its items are identical to
2461 those for 4.50. This seems to be the most sensible way to preserve the
2462 historical information.
2468 1. Minor wording change to the doc/README.SIEVE file.
2470 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2471 computation of the current number of files was incorrect.
2473 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2474 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2477 4. Give more explanation in the error message when the command for a transport
2478 filter fails to execute.
2480 5. There are several places where Exim runs a non-Exim command in a
2481 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2482 was being done only for the command run by the queryprogram router. It is
2483 now done for all such subprocesses. The other cases are: ${run, transport
2484 filters, and the commands run by the lmtp and pipe transports.
2486 6. Added CONFIGURE_GROUP build-time option.
2488 7. Some older OS have a limit of 256 on the maximum number of file
2489 descriptors. Exim was using setrlimit() to set 1000 as a large value
2490 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2491 systems. I've change it so that if it can't get 1000, it tries for 256.
2493 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2494 was an oversight, and furthermore, ever since the addition of extra
2495 controls (e.g. 4.43/32), the checks on when to allow different forms of
2496 "control" were broken. There should now be diagnostics for all cases when a
2497 control that does not make sense is encountered.
2499 9. Added the /retain_sender option to "control=submission".
2501 10. $recipients is now available in the predata ACL (oversight).
2503 11. Tidy the search cache before the fork to do a delivery from a message
2504 received from the command line. Otherwise the child will trigger a lookup
2505 failure and thereby defer the delivery if it tries to use (for example) a
2506 cached ldap connection that the parent has called unbind on.
2508 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2509 of $address_data from the recipient verification was clobbered by the
2510 sender verification.
2512 13. The value of address_data from a sender verification is now available in
2513 $sender_address_data in subsequent conditions in the ACL statement.
2515 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2517 15. Added a new option "connect=<time>" to callout options, to set a different
2520 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2521 was its contents. (It was OK if the option was not defined at all.)
2523 17. A "Completed" log line is now written for messages that are removed from
2524 the spool by the -Mrm option.
2526 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2527 information about exactly what failed.
2529 19. Added -dd to debug only the daemon process.
2531 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2532 handles timeouts, both on the server side and network timeouts. Renamed the
2533 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2535 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2536 It is now set to "smtps".
2538 22. $host_address is now set to the target address during the checking of
2539 ignore_target_hosts.
2541 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2542 being passed; this would have caused $sender_host_name to have been used if
2543 matching the list had actually called for a host name (not very likely,
2544 since this list is usually IP addresses). A host name is now passed as
2547 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2548 code that specifies a non-restarting handler (typically sigaction() in
2549 modern systems) in an attempt to fix a rare and obscure crash bug.
2551 25. Narrowed the window for a race in the daemon that could cause it to ignore
2552 SIGCHLD signals. This is not a major problem, because they are used only to
2553 wake it up if nothing else does.
2555 26. A malformed maildirsize file could cause Exim to calculate negative values
2556 for the mailbox size or file count. Odd effects could occur as a result.
2557 The maildirsize information is now recalculated if the size or filecount
2560 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2561 support for a long time. Removed HAVE_SYS_VFS_H.
2563 28. Installed the latest version of exipick from John Jetmore.
2565 29. In an address list, if the pattern was not a regular expression, an empty
2566 subject address (from a bounce message) matched only if the pattern was an
2567 empty string. Non-empty patterns were not even tested. This was the wrong
2568 because it is perfectly reasonable to use an empty address as part of a
2569 database query. An empty address is now tested by patterns that are
2570 lookups. However, all the other forms of pattern expect the subject to
2571 contain a local part and a domain, and therefore, for them, an empty
2572 address still always fails if the pattern is not itself empty.
2574 30. Exim went into a mad DNS loop when attempting to do a callout where the
2575 host was specified on an smtp transport, and looking it up yielded more
2576 than one IP address.
2578 31. Re-factored the code for checking spool and log partition space into a
2579 function that finds that data and another that does the check. The former
2580 is then used to implement four new variables: $spool_space, $log_space,
2581 $spool_inodes, and $log_inodes.
2583 32. The RFC2047 encoding function was originally intended for short strings
2584 such as real names; it was not keeping to the 75-character limit for
2585 encoded words that the RFC imposes. It now respects the limit, and
2586 generates multiple encoded words if necessary. To be on the safe side, I
2587 have increased the buffer size for the ${rfc2047: expansion operator from
2590 33. It is now permitted to omit both strings after an "if" condition; if the
2591 condition is true, the result is "true". As before, when the second string
2592 is omitted, a false condition yields an empty string. This makes it less
2593 cumbersome to write custom ACL and router conditions.
2595 34. Failure to deliver a bounce message always caused it to be frozen, even if
2596 there was an errors_to setting on the router. The errors_to setting is now
2599 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2600 canonical form (fully expanded) before being placed in
2601 $sender_host_address.
2603 36. The table in the code that translates DNS record types into text (T_A to
2604 "A" for instance) was missing entries for NS and CNAME. It is just possible
2605 that this could have caused confusion if both these types were looked up
2606 for the same domain, because the text type is used as part of Exim's
2607 per-process caching. But the chance of anyone hitting this buglet seems
2610 37. The dnsdb lookup has been extended in a number of ways.
2612 (1) There is a new type, "zns", which walks up the domain tree until it
2613 finds some nameserver records. It should be used with care.
2615 (2) There is a new type, "mxh", which is like "mx" except that it returns
2616 just the host names, not the priorities.
2618 (3) It is now possible to give a list of domains (or IP addresses) to be
2619 looked up. The behaviour when one of the lookups defers can be
2620 controlled by a keyword.
2622 (4) It is now possible to specify the separator character for use when
2623 multiple records are returned.
2625 38. The dnslists ACL condition has been extended: it is now possible to supply
2626 a list of IP addresses and/or domains to be looked up in a particular DNS
2629 39. Added log_selector=+queue_time_overall.
2631 40. When running the queue in the test harness, wait just a tad after forking a
2632 delivery process, to get repeatability of debugging output.
2634 41. Include certificate and key file names in error message when GnuTLS fails
2635 to set them up, because the GnuTLS error message doesn't include the name
2636 of the failing file when there is a problem reading it.
2638 42. Allow both -bf and -bF in the same test run.
2640 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2642 44. The "Exiscan patch" is now merged into the mainline Exim source.
2644 45. Sometimes the final signoff response after QUIT could fail to get
2645 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2646 < 0 before doing a fflush(). This bug looks as though it goes back to the
2647 introduction of TLS in release 3.20, but "sometimes" must have been rare
2648 because the tests only now provoked it.
2650 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2651 (this can affect the format of dates).
2653 47. exim_tidydb, when checking for the continued existence of a message for
2654 which it has found a message-specific retry record, was not finding
2655 messages that were in split spool directories. Consequently, it was
2656 deleting retry records that should have stayed in existence.
2658 48. Steve fixed some bugs in eximstats.
2660 49. The SPA authentication driver was not abandoning authentication and moving
2661 on to the next authenticator when an expansion was forced to fail,
2662 contradicting the general specification for all authenticators. Instead it
2663 was generating a temporary error. It now behaves as specified.
2665 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2666 (the order specifies the preference for clients). The order is now AES256,
2667 AES128, 3DES, ARCFOUR128.
2669 51. Small patch to Sieve code - explicitly set From: when generating an
2672 52. Exim crashed if a remote delivery caused a very long error message to be
2673 recorded - for instance if somebody sent an entire SpamAssassin report back
2674 as a large number of 550 error lines. This bug was coincidentally fixed by
2675 increasing the size of one of Exim's internal buffers (big_buffer) that
2676 happened as part of the Exiscan merge. However, to be on the safe side, I
2677 have made the code more robust (and fixed the comments that describe what
2680 53. Now that there can be additional text after "Completed" in log lines (if
2681 the queue_time_overall log selector is set), a one-byte patch to exigrep
2682 was needed to allow it to recognize "Completed" as not the last thing in
2685 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2686 patch that reportedly fixes this has been added. I am not expert enough to
2687 create a test for it. This is what the patch creator wrote:
2689 "I found a little strange behaviour of ldap code when working with
2690 Windows 2003 AD Domain, where users was placed in more than one
2691 Organization Units. When I tried to give exim partial DN, the exit code
2692 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2693 But simultaneously result of request was absolutely normal ldap result,
2694 so I produce this patch..."
2696 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2697 REFERENCE, so I have modified the code to exclude the patch when that macro
2700 55. Some experimental protocols are using DNS PTR records for new purposes. The
2701 keys for these records are domain names, not reversed IP addresses. The
2702 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2703 leaves it alone. Component reversal etc. now happens only for IP addresses.
2706 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2708 57. Double the size of the debug message buffer (to 2048) so that more of very
2709 long debug lines gets shown.
2711 58. The exicyclog utility now does better if the number of log files to keep
2712 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2714 59. Two changes related to the smtp_active_hostname option:
2716 (1) $smtp_active_hostname is now available as a variable.
2717 (2) The default for smtp_banner uses $smtp_active_hostname instead
2718 of $primary_hostname.
2720 60. The host_aton() function is supposed to be passed a string that is known
2721 to be a valid IP address. However, in the case of IPv6 addresses, it was
2722 not checking this. This is a hostage to fortune. Exim now panics and dies
2723 if the condition is not met. A case was found where this could be provoked
2724 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2725 components; fortuitously, this particular loophole had already been fixed
2726 by change 4.50/55 above.
2728 If there are any other similar loopholes, the new check in host_aton()
2729 itself should stop them being exploited. The report I received stated that
2730 data on the command line could provoke the exploit when Exim was running as
2731 exim, but did not say which command line option was involved. All I could
2732 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2733 running as the user.
2736 61. There was a buffer overflow vulnerability in the SPA authentication code
2737 (which came originally from the Samba project). I have added a test to the
2738 spa_base64_to_bits() function which I hope fixes it.
2741 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2742 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2744 63. The daemon start-up calls getloadavg() while still root for those OS that
2745 need the first call to be done as root, but it missed one case: when
2746 deliver_queue_load_max is set with deliver_drop_privilege. This is
2747 necessary for the benefit of the queue runner, because there is no re-exec
2748 when deliver_drop_privilege is set.
2750 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2751 This has been fixed.
2753 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2754 was in use, was not putting the data itself into the right store pool;
2755 consequently, it could be overwritten for a subsequent message in the same
2756 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2759 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2761 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2762 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2763 the function were treating the return as a boolean value, which happened to
2764 work because 0=false and not-0=true, but is not correct code.
2766 68. The host_aton() function was not handling scoped IPv6 addresses (those
2767 with, for example, "%eth0" on the end) correctly.
2769 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2770 negated items (that is, ~something) in unsigned ints. Some compilers
2771 apparently mutter when there is no cast.
2773 70. If an address verification called from an ACL failed, and did not produce a
2774 user-specific message (i.e. there was only a "system" message), nothing was
2775 put in $acl_verify_message. In this situation, it now puts the system
2778 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2779 session; change 4.31/43 added the unwanted input to the log line - except
2780 that it did not do this in the start of session case. It now does.
2782 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2783 This is wrong and can cause the other end to generate a synchronization
2784 error if it is another Exim or anything else that does the synchronization
2785 check. A QUIT command is no longer sent after a timeout.
2787 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2788 during host lookups.
2790 74. The defer_ok option of callout verification was not working if it was used
2791 when verifying addresses in header lines, that is, for this case:
2793 verify = header_sender/callout=defer_ok
2795 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2796 those file descriptors could be used for SMTP connections. If anything
2797 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2798 could be sent to the SMTP client, causing chaos. The daemon now opens
2799 stdin, stdout, and stderr to /dev/null when it puts itself into the
2802 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2803 log by default. The user can override this with suitable Perl magic.
2805 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2806 the log message when discard triggers, was not working for the DATA ACL or
2807 for the non-SMTP ACL.
2809 78. Error message wording change in sieve.c.
2811 79. If smtp_accept_max_per_host was set, the number of connections could be
2812 restricted to fewer than expected, because the daemon was trying to set up
2813 a new connection before checking whether the processes handling previous
2814 connections had finished. The check for completed processes is now done
2815 earlier. On busy systems, this bug wouldn't be noticed because something
2816 else would have woken the daemon, and it would have reaped the completed
2819 80. If a message was submitted locally by a user whose login name contained one
2820 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2821 It caused a spool format error. I have fixed the spool reading code. A
2822 related problem was that the "from" clause in the Received: line became
2823 illegal because of the space(s). It is now covered by ${quote_local_part.
2825 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2828 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2830 83. Patch from Sieve maintainer to fix unterminated string problem in
2831 "vacation" handling.
2833 84. Some minor changes to the Linux configuration files to help with other
2834 OS variants using glibc.
2836 85. One more patch for Sieve to update vacation handling to latest spec.
2839 ----------------------------------------------------
2840 See the note above about the 4.44 and 4.50 releases.
2841 ----------------------------------------------------
2847 1. Change 4.43/35 introduced a bug that caused file counts to be
2848 incorrectly computed when quota_filecount was set in an appendfile
2851 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2852 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2855 3. Give more explanation in the error message when the command for a transport
2856 filter fails to execute.
2858 4. There are several places where Exim runs a non-Exim command in a
2859 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2860 was being done only for the command run by the queryprogram router. It is
2861 now done for all such subprocesses. The other cases are: ${run, transport
2862 filters, and the commands run by the lmtp and pipe transports.
2864 5. Some older OS have a limit of 256 on the maximum number of file
2865 descriptors. Exim was using setrlimit() to set 1000 as a large value
2866 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2867 systems. I've change it so that if it can't get 1000, it tries for 256.
2869 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2870 was an oversight, and furthermore, ever since the addition of extra
2871 controls (e.g. 4.43/32), the checks on when to allow different forms of
2872 "control" were broken. There should now be diagnostics for all cases when a
2873 control that does not make sense is encountered.
2875 7. $recipients is now available in the predata ACL (oversight).
2877 8. Tidy the search cache before the fork to do a delivery from a message
2878 received from the command line. Otherwise the child will trigger a lookup
2879 failure and thereby defer the delivery if it tries to use (for example) a
2880 cached ldap connection that the parent has called unbind on.
2882 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2883 of $address_data from the recipient verification was clobbered by the
2884 sender verification.
2886 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2887 was its contents. (It was OK if the option was not defined at all.)
2889 11. A "Completed" log line is now written for messages that are removed from
2890 the spool by the -Mrm option.
2892 12. $host_address is now set to the target address during the checking of
2893 ignore_target_hosts.
2895 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2896 being passed; this would have caused $sender_host_name to have been used if
2897 matching the list had actually called for a host name (not very likely,
2898 since this list is usually IP addresses). A host name is now passed as
2901 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2902 code that specifies a non-restarting handler (typically sigaction() in
2903 modern systems) in an attempt to fix a rare and obscure crash bug.
2905 15. Narrowed the window for a race in the daemon that could cause it to ignore
2906 SIGCHLD signals. This is not a major problem, because they are used only to
2907 wake it up if nothing else does.
2909 16. A malformed maildirsize file could cause Exim to calculate negative values
2910 for the mailbox size or file count. Odd effects could occur as a result.
2911 The maildirsize information is now recalculated if the size or filecount
2914 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2915 support for a long time. Removed HAVE_SYS_VFS_H.
2917 18. Updated exipick to current release from John Jetmore.
2919 19. Allow an empty sender to be matched against a lookup in an address list.
2920 Previously the only cases considered were a regular expression, or an
2923 20. Exim went into a mad DNS lookup loop when doing a callout where the
2924 host was specified on the transport, if the DNS lookup yielded more than
2927 21. The RFC2047 encoding function was originally intended for short strings
2928 such as real names; it was not keeping to the 75-character limit for
2929 encoded words that the RFC imposes. It now respects the limit, and
2930 generates multiple encoded words if necessary. To be on the safe side, I
2931 have increased the buffer size for the ${rfc2047: expansion operator from
2934 22. Failure to deliver a bounce message always caused it to be frozen, even if
2935 there was an errors_to setting on the router. The errors_to setting is now
2938 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2939 canonical form (fully expanded) before being placed in
2940 $sender_host_address.
2942 24. Updated eximstats to version 1.33
2944 25. Include certificate and key file names in error message when GnuTLS fails
2945 to set them up, because the GnuTLS error message doesn't include the name
2946 of the failing file when there is a problem reading it.
2948 26. Expand error message when OpenSSL has problems setting up cert/key files.
2951 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2952 (this can affect the format of dates).
2954 28. exim_tidydb, when checking for the continued existence of a message for
2955 which it has found a message-specific retry record, was not finding
2956 messages that were in split spool directories. Consequently, it was
2957 deleting retry records that should have stayed in existence.
2959 29. eximstats updated to version 1.35
2960 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2961 1.35 - bugfix such that pie charts by volume are generated correctly
2963 30. The SPA authentication driver was not abandoning authentication and moving
2964 on to the next authenticator when an expansion was forced to fail,
2965 contradicting the general specification for all authenticators. Instead it
2966 was generating a temporary error. It now behaves as specified.
2968 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2969 (the order specifies the preference for clients). The order is now AES256,
2970 AES128, 3DES, ARCFOUR128.
2972 31. Small patch to Sieve code - explicitly set From: when generating an
2975 32. Exim crashed if a remote delivery caused a very long error message to be
2976 recorded - for instance if somebody sent an entire SpamAssassin report back
2977 as a large number of 550 error lines. This bug was coincidentally fixed by
2978 increasing the size of one of Exim's internal buffers (big_buffer) that
2979 happened as part of the Exiscan merge. However, to be on the safe side, I
2980 have made the code more robust (and fixed the comments that describe what
2983 33. Some experimental protocols are using DNS PTR records for new purposes. The
2984 keys for these records are domain names, not reversed IP addresses. The
2985 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2986 leaves it alone. Component reversal etc. now happens only for IP addresses.
2989 34. The host_aton() function is supposed to be passed a string that is known
2990 to be a valid IP address. However, in the case of IPv6 addresses, it was
2991 not checking this. This is a hostage to fortune. Exim now panics and dies
2992 if the condition is not met. A case was found where this could be provoked
2993 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2994 components; fortuitously, this particular loophole had already been fixed
2995 by change 4.50/55 or 4.44/33 above.
2997 If there are any other similar loopholes, the new check in host_aton()
2998 itself should stop them being exploited. The report I received stated that
2999 data on the command line could provoke the exploit when Exim was running as
3000 exim, but did not say which command line option was involved. All I could
3001 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3002 running as the user.
3005 35. There was a buffer overflow vulnerability in the SPA authentication code
3006 (which came originally from the Samba project). I have added a test to the
3007 spa_base64_to_bits() function which I hope fixes it.
3010 36. The daemon start-up calls getloadavg() while still root for those OS that
3011 need the first call to be done as root, but it missed one case: when
3012 deliver_queue_load_max is set with deliver_drop_privilege. This is
3013 necessary for the benefit of the queue runner, because there is no re-exec
3014 when deliver_drop_privilege is set.
3016 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3017 was in use, was not putting the data itself into the right store pool;
3018 consequently, it could be overwritten for a subsequent message in the same
3019 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3022 38. Sometimes the final signoff response after QUIT could fail to get
3023 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3024 < 0 before doing a fflush(). This bug looks as though it goes back to the
3025 introduction of TLS in release 3.20, but "sometimes" must have been rare
3026 because the tests only now provoked it.
3032 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3033 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3034 Now it returns an integer. A number of calls were still expecting a T/F
3035 return. Fortuitously, in all cases, the tests worked in OK situations,
3036 which is the norm. However, things would have gone wrong on any write
3037 failures on the smtp file descriptor. This function is used when sending
3038 messages over SMTP and also when doing verify callouts.
3040 2. When Exim is called to do synchronous delivery of a locally submitted
3041 message (the -odf or -odi options), it no longer closes stderr before doing
3044 3. Implemented the mua_wrapper option.
3046 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3048 5. Implemented the functions header_remove(), header_testname(),
3049 header_add_at_position(), and receive_remove_recipient(), and exported them
3052 6. If an ACL "warn" statement specified the addition of headers, Exim already
3053 inserted X-ACL-Warn: at the start if there was no header name. However, it
3054 was not making this test for the second and subsequent header lines if
3055 there were newlines in the string. This meant that an invalid header could
3056 be inserted if Exim was badly configured.
3058 7. Allow an ACL "warn" statement to add header lines at the start or after all
3059 the Received: headers, as well as at the end.
3061 8. Added the rcpt_4xx retry error code.
3063 9. Added postmaster_mailfrom=xxx to callout verification option.
3065 10. Added mailfrom=xxxx to the callout verification option, for verify=
3068 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3069 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3071 12. Inserted some casts to stop certain compilers warning when using pointer
3072 differences as field lengths or precisions in printf-type calls (mostly
3073 affecting debugging statements).
3075 13. Added optional readline() support for -be (dynamically loaded).
3077 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3078 same clock tick as a message's arrival, so that its received time was the
3079 same as the "first fail" time on the retry record, and that message
3080 remained on the queue past the ultimate address timeout, every queue runner
3081 would try a delivery (because it was past the ultimate address timeout) but
3082 after another failure, the ultimate address timeout, which should have then
3083 bounced the address, did not kick in. This was a "< instead of <=" error;
3084 in most cases the first failure would have been in the next clock tick
3085 after the received time, and all would be well.
3087 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3088 being recognized when the domain list was tested by the match_domain
3089 condition in an expansion string.
3091 16. Added the ${str2b64: operator.
3093 17. Exim was always calling setrlimit() to set a large limit for the number of
3094 processes, without checking whether the existing limit was already
3095 adequate. (It did check for the limit on file descriptors.) Furthermore,
3096 errors from getrlimit() and setrlimit() were being ignored. Now they are
3097 logged to the main and panic logs, but Exim does carry on, to try to do its
3098 job under whatever limits there are.
3100 18. Imported PCRE 5.0.
3102 19. Trivial typo in log message " temporarily refused connection" (the leading
3105 20. If the log selector return_path_on_delivery was set and an address was
3106 redirected to /dev/null, the delivery process crashed because it assumed
3107 that a return path would always be set for a "successful" delivery. In this
3108 case, the whole delivery is bypassed as an optimization, and therefore no
3111 21. Internal re-arrangement: the function for sending a challenge and reading
3112 a response while authentication was assuming a zero-terminated challenge
3113 string. It's now changed to take a pointer and a length, to allow for
3114 binary data in such strings.
3116 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3118 23. Exim was not respecting finduser_retries when seeking the login of the
3119 uid under which it was called; it was always trying 10 times. (The default
3120 setting of finduser_retries is zero.) Also, it was sleeping after the final
3121 failure, which is pointless.
3123 24. Implemented tls_on_connect_ports.
3125 25. Implemented acl_smtp_predata.
3127 26. If the domain in control=submission is set empty, Exim assumes that the
3128 authenticated id is a complete email address when it generates From: or
3129 Sender: header lines.
3131 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3132 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3133 chown and chgrp in /bin and hostname in /usr/bin.
3135 28. Exim was keeping the "process log" file open after each use, just as it
3136 does for the main log. This opens the possibility of it remaining open for
3137 long periods when the USR1 signal hits a daemon. Occasional processlog
3138 errors were reported, that could have been caused by this. Anyway, it seems
3139 much more sensible not to leave this file open at all, so that is what now
3142 29. The long-running daemon process does not normally write to the log once it
3143 has entered its main loop, and it closes the log before doing so. This is
3144 so that log files can straightforwardly be renamed and moved. However,
3145 there are a couple of unusual error situations where the daemon does write
3146 log entries, and I had neglected to close the log afterwards.
3148 30. The text of an SMTP error response that was received during a remote
3149 delivery was being truncated at 512 bytes. This is too short for some of
3150 the long messages that one sometimes sees. I've increased the limit to
3153 31. It is now possible to make retry rules that apply only when a message has a
3154 specific sender, in particular, an empty sender.
3156 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3157 it possible to be selective about when SMTP synchronization is enforced.
3159 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3161 32. Implemented hosts_connection_nolog.
3163 33. Added an ACL for QUIT.
3165 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3168 35. Added mailbox_size and mailbox_filecount to appendfile.
3170 36. Added control = no_multiline_responses to ACLs.
3172 37. There was a bug in the logic of the code that waits for the clock to tick
3173 in the case where the clock went backwards by a substantial amount such
3174 that the microsecond fraction of "now" was more than the microsecond
3175 fraction of "then" (but the whole seconds number was less).
3177 38. Added support for the libradius Radius client library this is found on
3178 FreeBSD (previously only the radiusclient library was supported).
3184 1. When certain lookups returned multiple values in the form name=value, the
3185 quoting of the values was not always being done properly. Specifically:
3186 (a) If the value started with a double quote, but contained no whitespace,
3188 (b) If the value contained whitespace other than a space character (i.e.
3189 tabs or newlines or carriage returns) it was not quoted.
3190 This fix has been applied to the mysql and pgsql lookups by writing a
3191 separate quoting function and calling it from the lookup code. The fix
3192 should probably also be applied to nisplus, ibase and oracle lookups, but
3193 since I cannot test any of those, I have not disturbed their existing code.
3195 2. A hit in the callout cache for a specific address caused a log line with no
3196 reason for rejecting RCPT. Now it says "Previous (cached) callout
3197 verification failure".
3199 3. There was an off-by-one bug in the queryprogram router. An over-long
3200 return line was truncated at 256 instead of 255 characters, thereby
3201 overflowing its buffer with the terminating zero. As well as fixing this, I
3202 have increased the buffer size to 1024 (and made a note to document this).
3204 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3205 when Exim is waiting for an SMTP response from a remote server, Exim
3206 restarts its select() call on the socket, thereby resetting its timeout.
3207 This is not a problem when such interrupts are rare. Somebody set up a cron
3208 job to run exiwhat every 2 minutes, which is less than the normal select()
3209 timeout (5 or 10 minutes). This meant that the select() timeout never
3210 kicked in because it was always reset. I have fixed this by comparing the
3211 time when an interrupt arrives with the time at the start of the first call
3212 to select(). If more time than the timeout has elapsed, the interrupt is
3213 treated as a timeout.
3215 5. Some internal re-factoring in preparation for the addition of Sieve
3216 extensions (by MH). In particular, the "personal" test is moved to a
3217 separate function, and given an option for scanning Cc: and Bcc: (which is
3218 not set for Exim filters).
3220 6. When Exim created an email address using the login of the caller as the
3221 local part (e.g. when creating a From: or Sender: header line), it was not
3222 quoting the local part when it contained special characters such as @.
3224 7. Installed new OpenBSD configuration files.
3226 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3227 try to make them clearer.
3229 9. Callout options, other than the timeout value, were being ignored when
3230 verifying sender addresses in header lines. For example, when using
3232 verify = header_sender/callout=no_cache
3234 the cache was (incorrectly) being used.
3236 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3237 only the Cygwin environment.
3239 11. When return_path_on_delivery was set as a log selector, if different remote
3240 addresses in the same message used different return paths and parallel
3241 remote delivery occurred, the wrong values would sometimes be logged.
3242 (Whenever a remote delivery process finished, the return path value from
3243 the most recently started remote delivery process was logged.)
3245 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3246 lines when AUTH and/or TLS are in use. This is the "received protocol"
3247 field. Exim used to use "asmtp" for authenticated SMTP, without any
3248 indication (in the protocol name) for TLS use. Now it follows the RFC and
3249 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3250 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3251 names appear in log lines as well as in Received: header lines.
3253 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3254 extensions, and comparison tests, and to fix some bugs.
3256 14. Changes to the "personal" filter test:
3258 (1) The test was buggy in that it was just doing the equivalent of
3259 "contains" tests on header lines. For example, if a user's address was
3260 anne@some.where, the "personal" test would incorrectly be true for
3262 To: susanne@some.where
3264 This test is now done by extracting each address from the header in turn,
3265 and checking the entire address. Other tests that are part of "personal"
3266 are now done using regular expressions (for example, to check local parts
3267 of addresses in From: header lines).
3269 (2) The list of non-personal local parts in From: addresses has been
3270 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3271 taken from the Sieve specification recommendations.
3273 (3) If the message contains any header line starting with "List-" it is
3274 treated as non-personal.
3276 (4) The test for "circular" in the Subject: header line has been removed
3277 because it now seems ill-conceived.
3279 15. Minor typos in src/EDITME comments corrected.
3281 16. Installed latest exipick from John Jetmore.
3283 17. If headers_add on a router specified a text string that was too long for
3284 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3285 of string_sprintf() is now avoided.
3287 18. $message_body_size was not set (it was always zero) when running the DATA
3288 ACL and the local_scan() function.
3290 19. For the "mail" command in an Exim filter, no default was being set for
3291 the once_repeat time, causing a random time value to be used if "once" was
3292 specified. (If the value happened to be <= 0, no repeat happened.) The
3293 default is now 0s, meaning "never repeat". The "vacation" command was OK
3294 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3295 (I found it when inspecting the code).
3297 20. There is now an overall timeout for performing a callout verification. It
3298 defaults to 4 times the callout timeout, which applies to individual SMTP
3299 commands during the callout. The overall timeout applies when there is more
3300 than one host that can be tried. The timeout is checked before trying the
3301 next host. This prevents very long delays if there are a large number of
3302 hosts and all are timing out (e.g. when the network connections are timing
3303 out). The value of the overall timeout can be changed by specifying an
3304 additional sub-option for "callout", called "maxwait". For example:
3306 verify = sender/callout=5s,maxwait=20s
3308 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3309 to the end before writing, but this should make it even safer).
3311 22. Exim was forgetting that it had advertised PIPELINING for the second and
3312 subsequent messages on an SMTP connection. It was also not resetting its
3313 memory on STARTTLS and an internal HELO.
3315 23. When Exim logs an SMTP synchronization error within a session, it now
3316 records whether PIPELINING has been advertised or not.
3318 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3319 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3320 rather than long int.
3322 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3324 26. Added the never_mail option to autoreply.
3330 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3331 crash if the getsockname() call failed; this can happen if a connection is
3332 closed very soon after it is established. The problem was simply in the
3333 order in which certain operations were done, causing Exim to try to write
3334 to the SMTP stream before it had set up the file descriptor. The bug has
3335 been fixed by making things happen in the correct order.
3341 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3342 before the connection was closed, thus losing the rejection response.
3344 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3345 some early Solaris releases, but causes trouble in current releases where
3346 socklen_t is defined.
3348 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3351 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3354 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3355 that releasing the top of it at the end releases what was used for sub-
3356 expansions (unless the block got too big). However, discard this block if
3357 the first thing is a variable or header, so that we can use its block when
3358 it is dynamic (useful for very large $message_headers, for example).
3360 6. Lookups now cache *every* query, not just the most recent. A new, separate
3361 store pool is used for this. It can be recovered when all lookup caches are
3362 flushed. Lookups now release memory at the end of their result strings.
3363 This has involved some general refactoring of the lookup sources.
3365 7. Some code has been added to the store_xxx() functions to reduce the amount
3366 of flapping under certain conditions.
3368 8. log_incoming_interface used to affect only the <= reception log lines. Now
3369 it causes the local interface and port to be added to several more SMTP log
3370 lines, for example "SMTP connection from", and rejection lines.
3372 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3374 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3376 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3377 could be overwritten at the end of the current message (or the start of a
3378 new message if it was set in a HELO ACL). The value is now preserved for
3379 the duration of the SMTP connection.
3381 12. If a transport had a headers_rewrite setting, and a matching header line
3382 contained an unqualified address, that address was qualified, even if it
3383 did not match any rewriting rules. The underlying bug was that the values
3384 of the flags that permit the existence of unqualified sender and recipient
3385 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3386 non-local messages, and by -bnq for local messages) were not being
3387 preserved with the message after it was received.
3389 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3390 "next input=" as part of the text comprising the host identity instead of
3391 the correct text. The code was using the same buffer for two different
3392 strings. However, depending on which order the printing function evaluated
3393 its arguments, the bug did not always show up. Under Linux, for example, my
3394 test suite worked just fine.
3396 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3397 doesn't work with some older versions of Perl. It has been changed to "my",
3398 which in any case is probably the better facility to use.
3400 15. A really picky compiler found some instances of statements for creating
3401 error messages that either had too many or two few arguments for the format
3404 16. The size of the buffer for calls to the DNS resolver has been increased
3405 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3406 for addresses that have a lot of PTR records. This alleviates a problem; it
3407 does not fully solve it.
3409 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3410 buffer now truncates the list and logs the incident, which is the same
3411 action as happens when Exim is looking up a host name and its aliases.
3412 Previously in this situation something unpredictable would happen;
3413 sometimes it was "internal error: store_reset failed".
3415 18. If a server dropped the connection unexpectedly when an Exim client was
3416 using GnuTLS and trying to read a response, the client delivery process
3417 crashed while trying to generate an error log message.
3419 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3420 string, for example:
3422 warn message = H1: something\nH2: something
3424 the text was added as a single header line from Exim's point of view
3425 though it ended up OK in the delivered message. However, searching for the
3426 second and subsequent header lines using $h_h2: did not work. This has been
3427 fixed. Similarly, if a system filter added multiple headers in this way,
3428 the routers could not see them.
3430 20. Expanded the error message when iplsearch is called with an invalid key to
3431 suggest using net-iplsearch in a host list.
3433 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3434 ACLs are no longer actually imposed (and a message to that effect is
3437 22. If a "gecos" field in a passwd entry contained escaped characters, in
3438 particular, if it contained a \" sequence, Exim got it wrong when building
3439 a From: or a Sender: header from that name. A second bug also caused
3440 incorrect handling when an unquoted " was present following a character
3441 that needed quoting.
3443 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3444 was not being matched caselessly.
3446 24. Arranged for all hyphens in the exim.8 source to be escaped with
3449 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3450 properly. Recipient callout cache records were still being keyed to include
3451 the sender, even when use_sender was set false. This led to far more
3452 callouts that were necessary. The sender is no longer included in the key
3453 when use_sender is false.
3455 26. Added "control = submission" modifier to ACLs.
3457 27. Added the ${base62d: operator to decode base 62 numbers.
3459 28. dnsdb lookups can now access SRV records.
3461 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3462 the configuration file.
3464 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3465 (-v) mode. This makes the output for a verbose queue run more intelligible.
3467 31. Added a use_postmaster feature to recipient callouts.
3469 32. Added the $body_zerocount variable, containing the number of binary zero
3470 bytes in the message body.
3472 33. The time of last modification of the "new" subdirectory is now used as the
3473 "mailbox time last read" when there is a quota error for a maildir
3476 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3478 35. Added +ignore_unknown as a special item in host lists.
3480 36. Code for decoding IPv6 addresses in host lists is now included, even if
3481 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3482 address was recognized as an IP address, but was then not correctly decoded
3483 into binary, causing unexpected and incorrect effects when compared with
3490 1. Very minor rewording of debugging text in manualroute to say "list of
3491 hosts" instead of "hostlist".
3493 2. If verify=header_syntax was set, and a header line with an unqualified
3494 address (no domain) and a large number of spaces between the end of the
3495 name and the colon was received, the reception process suffered a buffer
3496 overflow, and (when I tested it) crashed. This was caused by some obsolete
3497 code that should have been removed. The fix is to remove it!
3499 3. When running in the test harness, delay a bit after writing a bounce
3500 message to get a bit more predictability in the log output.
3502 4. Added a call to search_tidyup() just before forking a reception process. In
3503 theory, someone could use a lookup in the expansion of smtp_accept_max_
3504 per_host which, without the tidyup, could leave open a database connection.
3506 5. Added the variables $recipient_data and $sender_data which get set from a
3507 lookup success in an ACL "recipients" or "senders" condition, or a router
3508 "senders" option, similar to $domain_data and $local_part_data.
3510 6. Moved the writing of debug_print from before to after the "senders" test
3513 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3514 problems for message scanning, either using a data ACL, or using
3515 local_scan() because the Received: header was not generated till after they
3516 were called (in order to set the time as the time of reception completion).
3517 I have revised the way this works. The header is now generated after the
3518 body is received, but before the ACL or local_scan() are called. After they
3519 are run, the timestamp in the header is updated.
3525 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3526 before starting a queue runner without re-exec. This happened only when
3527 deliver_drop_privilege was set or when the Exim user was set to root. The
3528 effect of the bug was that timeouts during subsequent deliveries caused
3529 crashes instead of being properly handled. The handler is now left at its
3530 default (and expected) setting.
3532 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3533 message, again when deliver_drop_privilege is set or Exim is run as root.
3534 The bug described in (1) was not present in this case, but the tidying up
3535 of the other signals was missing. I have made the two cases consistent.
3537 3. The ignore_target_hosts setting on a manualroute router was being ignored
3538 for hosts that were looked up using the /MX notation.
3540 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3543 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3544 operated on the sender address. After changing the $sender_address to <>
3545 for the sender address verify, Exim was re-instated it as the original
3546 (before rewriting) address, but remembering that it had rewritten it, so it
3547 wasn't rewriting it again. This bug also had the effect of breaking the
3548 sender address verification caching when the sender address was rewritten.
3550 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3551 This has been changed so that if the ip literal address matches
3552 ignore_target_hosts, the router declines.
3554 7. Added expansion conditions match_domain, match_address, and match_local_
3555 part (NOT match_host).
3557 8. The placeholder for the Received: header didn't have a length field set.
3559 9. Added code to Exim itself and to exim_lock to test for a specific race
3560 condition that could lead to file corruption when using MBX delivery. The
3561 issue is with the lockfile that is created in /tmp. If this file is removed
3562 after a process has opened it but before that process has acquired a lock,
3563 there is the potential for a second process to recreate the file and also
3564 acquire a lock. This could lead to two Exim processes writing to the file
3565 at the same time. The added code performs the same test as UW imapd; it
3566 checks after acquiring the lock that its file descriptor still refers to
3567 the same named file.
3569 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3570 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3573 11. Added the smtp_active_hostname option. If used, this will typically be made
3574 to depend on the incoming interface address. Because $interface_address is
3575 not set up until the daemon has forked a reception process, error responses
3576 that can happen earlier (such as "too many connections") no longer contain
3579 12. If an expansion in a condition on a "warn" statement fails because a lookup
3580 defers, the "warn" statement is abandoned, and the next ACL statement is
3581 processed. Previously this caused the whole ACL to be aborted.
3583 13. Added the iplsearch lookup type.
3585 14. Added ident_timeout as a log selector.
3587 15. Added tls_certificate_verified as a log selector.
3589 16. Added a global option tls_require_ciphers (compare the smtp transport
3590 option of the same name). This controls incoming TLS connections.
3592 17. I finally figured out how to make tls_require_ciphers do a similar thing
3593 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3594 before starting the TLS session.
3596 18. Tabs are now shown as \t in -bP output.
3598 19. If the log selector return_path_on_delivery was set, Exim crashed when
3599 bouncing a message because it had too many Received: header lines.
3601 20. If two routers both had headers_remove settings, and the first one included
3602 a superfluous trailing colon, the final name in the first list and the
3603 first name in the second list were incorrectly joined into one item (with a
3604 colon in the middle).
3610 1. Added -C and -D options to the exinext utility, mainly to make it easier
3611 to include in the automated testing, but these could be helpful when
3612 multiple configurations are in use.
3614 2. The exinext utility was not formatting the output nicely when there was
3615 an alternate port involved in the retry record key, nor when there was a
3616 message id as well (for retries that were specific to a specific message
3617 and a specific host). It was also confused by IPv6 addresses, because of
3618 the additional colons they contain. I have fixed the IPv4 problem, and
3619 patched it up to do a reasonable job for IPv6.
3621 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3622 delivery, the log line now contains "pipelined" if PIPELINING was used.
3624 4. An SMTP transport process used to panic and die if the bind() call to set
3625 an explicit outgoing interface failed. This has been changed; it is now
3626 treated in the same way as a connect() failure.
3628 5. A reference to $sender_host_name in the part of a conditional expansion
3629 that was being skipped was still causing a DNS lookup. This no longer
3632 6. The def: expansion condition was not recognizing references to header lines
3633 that used bh_ and bheader_.
3635 7. Added the _cache feature to named lists.
3637 8. The code for checking quota_filecount in the appendfile transport was
3638 allowing one more file than it should have been.
3640 9. For compatibility with Sendmail, the command line option
3648 and sets the incoming protocol and host name (for trusted callers). The
3649 host name and its colon can be omitted when only the protocol is to be set.
3650 Note the Exim already has two private options, -pd and -ps, that refer to
3651 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3652 "s", but I don't think that's a major issue.
3654 10. A number of refactoring changes to the code, none of which should affect
3657 (a) The number of logging options was getting close to filling up the
3658 32-bit word that was used as a bit map. I have split them into two classes:
3659 those that are passed in the argument to log_write(), and those that are
3660 only ever tested independently outside of that function. These are now in
3661 separate 32-bit words, so there is plenty of room for expansion again.
3662 There is no change in the user interface or the logging behaviour.
3664 (b) When building, for example, log lines, the code previously used a
3665 macro that called string_cat() twice, in order to add two strings. This is
3666 not really sufficiently general. Furthermore, there was one instance where
3667 it was actually wrong because one of the argument was used twice, and in
3668 one call a function was used. (As it happened, calling the function twice
3669 did not affect the overall behaviour.) The macro has been replaced by a
3670 function that can join an arbitrary number of extra strings onto a growing
3673 (c) The code for expansion conditions now uses a table and a binary chop
3674 instead of a serial search (which was left over from when there were very
3675 few conditions). Also, it now recognizes conditions like "pam" even when
3676 the relevant support is not compiled in: a suitably worded error message is
3677 given if an attempt is made to use such a condition.
3679 11. Added ${time_interval:xxxxx}.
3681 12. A bug was causing one of the ddress fields not to be passed back correctly
3682 from remote delivery subprocesses. The field in question was not being
3683 subsequently used, so this caused to problems in practice.
3685 13. Added new log selectors queue_time and deliver_time.
3687 14. Might have fixed a bug in maildirsizefile handling that threw up
3688 "unexpected character" debug warnings, and recalculated the data
3689 unnecessarily. In any case, I expanded the warning message to give more
3692 15. Added the message "Restricted characters in address" to the statements in
3693 the default ACL that block characters like @ and % in local parts.
3695 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3696 Three changes have been made:
3698 (a) There was a serious bug; a negative response to MAIL caused the whole
3699 recipient domain to be cached as invalid, thereby blocking all messages
3700 to all local parts at the same domain, from all senders. This bug has
3701 been fixed. The domain is no longer cached after a negative response to
3702 MAIL if the sender used is not empty.
3704 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3707 (c) A new callout option, "use_sender" has been added for people who want
3708 the modified behaviour.
3714 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3717 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3718 indeed breaks things for older releases.
3720 3. Added additional logging to the case where there is a problem reading data
3721 from a filter that is running in a subprocess using a pipe, in order to
3722 try to track down a specific problem.
3724 4. Testing facility fudge: when running in the test harness and attempting
3725 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3726 getting "No route to host". Convert this to a timeout.
3728 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3731 6. Some OS don't have socklen_t but use size_t instead. This affects the
3732 fifth argument of getsockopt() amongst other things. This is now
3733 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3734 can be set for individual OS. I have set it for SunOS5, OSF1, and
3735 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3736 some earlier ones do not.
3738 7. Change 4.30/15 was not doing the test caselessly.
3740 8. The standard form for an IPv6 address literal was being rejected by address
3741 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3742 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3743 this, as well as the form without the "IPv6" on the front (but only when
3744 address literals are enabled, of course).
3746 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3748 10. Exim crashed if a message with an empty sender address specified by -f
3749 encountered a router with an errors_to setting. This could be provoked only
3750 by a command such as
3754 where an empty string was supplied; "<>" did not hit this bug.
3756 11. Installed PCRE release 4.5.
3758 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3759 remained set. It is now erased.
3761 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3762 times from message ids (which are base 36 rather than the normal 62).
3764 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3765 were being counted as actual protocol errors, and logged if the log
3766 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3767 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3768 and DATA following a set of rejected RCPTs do not count as protocol errors.
3769 In other words, Exim assumes they were pipelined, though this may not
3770 actually be the case. Of course, in all cases the client gets an
3771 appropriate error code.
3773 15. If a lookup fails in an ACL condition, a message about the failure may
3774 be available; it is used if testing the ACL cannot continue, because most
3775 such messages specify what the cause of the deferral is. However, some
3776 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3777 that caused an old message to be retained and used if a later statement
3778 caused a defer, replacing the real cause of the deferral.
3780 16. If an IP address had so many PTR records that the DNS lookup buffer
3781 was not large enough to hold them, Exim could crash while trying to process
3782 the truncated data. It now detects and logs this case.
3784 17. Further to 4.21/58, another change has been made: if (and only if) the
3785 first line of a message (the first header line) ends with CRLF, a bare LF
3786 in a subsequent header line has a space inserted after it, so as not to
3787 terminate the header.
3789 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3790 unnecessarily, used atoi() instead of strtol(), and didn't check the
3791 termination when getting file sizes from file names by regex.
3793 19. Completely re-implemented the support for maildirsize files, in the light
3794 of a number of problems with the previous contributed implementation
3795 (4.30/29). In particular:
3797 . If the quota is zero, the maildirsize file is maintained, but no quota is
3800 . If the maildir directory does not exist, it is created before any attempt
3801 to write a maildirsize file.
3803 . The quota value in the file is just a cache; if the quota is changed in
3804 the transport, the new value overrides.
3806 . A regular expression is available for excluding directories from the
3809 20. The autoreply transport checks the characters in options that define the
3810 message's headers; it allows continued headers, but it was checking with
3811 isspace() after an embedded newline instead of explicitly looking for a
3814 21. If all the "regular" hosts to which an address was routed had passed their
3815 expiry times, and had not reached their retry times, the address was
3816 bounced, even if fallback hosts were defined. Now Exim should go on to try
3819 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3820 equivalent code in the SMTP transport. Some hosts send humungous responses
3821 to HELO/EHLO, more than 1024 it seems.
3823 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3824 gives compiler warnings in some environments. I've now done it "properly",
3827 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3828 (because of problems with the built-in one) was declared to return uschar *
3829 instead of char *, causing compiler failure.
3831 25. Fixed a file descriptor leak when processing alias/forward files.
3833 26. Fixed a minor format string issue in dbfn.c.
3835 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3837 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3838 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3841 29. When a sender address is verified, it is cached, to save repeating the test
3842 when there is more than one recipient in a message. However, when the
3843 verification involves a callout, it is possible for different callout
3844 options to be set for different recipients. It is too complicated to keep
3845 track of this in the cache, so now Exim always runs a verification when a
3846 callout is required, relying on the callout cache for the optimization.
3847 The overhead is duplication of the address routing, but this should not be
3850 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3851 to be verified with callout=postmaster, and the main callout worked but the
3852 postmaster check failed, the verification correctly failed. However, if a
3853 subsequent RCPT command asked for sender verification *without* the
3854 postmaster check, incorrect caching caused this verification also to fail,
3857 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3858 it was not caching the DNS options (qualify_single, search_parents) that
3859 were used when the lookup failed. A subsequent lookup with different
3860 options therefore always gave the same answer, though there were cases
3861 where it should not have. (Example: a "domains = !$mx_any" option on a
3862 dnslookup router: the "domains" option is always processed without any
3863 widening, but the router might have qualify_single set.) Now Exim uses the
3864 cached value only when the same options are set.
3866 32. Added John Jetmore's "exipick" utility to the distribution.
3868 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3869 than a timeout (e.g. a certificate is required, and is not provided), an
3870 Exim server now closes the connection immediately. Previously it waited for
3871 the client to close - but if the client is SSL, it seems that they each
3872 wait for each other, leading to a delay before one of them times out.
3874 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3875 maintained 0.8.x compatibility because I don't think many are using it, and
3876 it is clearly obsolete.
3878 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3881 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3882 client certificate was expired. A simple patch fixes this, though I don't
3883 understand the full logic of why the verify callback is called multiple
3886 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3887 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3888 which causes problems with some clients (such as the Certicom SSL Plus
3889 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3890 disables the coutermeasure allowing Eudora to connect."
3892 38. Exim was not checking that a write() to a log file succeeded. This could
3893 lead to Bad Things if a log got too big, in particular if it hit a file
3894 size limit. Exim now panics and dies if it cannot write to a log file, just
3895 as it does if it cannot open a log file.
3897 39. Modified OS/Makefile-Linux so that it now contains
3899 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3901 The two -D definitions ensure that Exim is compiled with large file
3902 support, which makes it possible to handle log files that are bigger than
3905 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3906 instance) a domain was checked against a named list that involved a lookup,
3907 causing $domain_data to be set, then another domain was checked against the
3908 same list, then the first domain was re-checked, the value of $domain_data
3909 after the final check could be wrong. In particular, if the second check
3910 failed, it could be set empty. This bug probably also applied to
3913 41. The strip_trailing_dot option was not being applied to the address given
3914 with the -f command-line option.
3916 42. The code for reading a message's header from the spool was incrementing
3917 $received_count, but never initializing it. This meant that the value was
3918 incorrect (doubled) while delivering a message in the same process in which
3919 it was received. In the most common configuration of Exim, this never
3920 happens - a fresh exec is done - but it can happen when
3921 deliver_drop_privilege is set.
3923 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3924 it now includes up to 150 characters of the unexpected data in the log
3927 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3928 and building data strings. The size of both of these buffers was 10 000
3929 bytes - far larger than anybody would *ever* want, thought I. Needless to
3930 say, somebody hit the limit. I have increased the maximum line length to
3931 20 000 and the maximum data length of concatenated lines to 100 000. I have
3932 also fixed two bugs, because there was no checking on these buffers. Tsk,
3933 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3934 buffer is too small.
3936 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3937 lsearch lookups. Now it does.
3939 46. When parsing a route_list item in a manualroute router, a fixed-length
3940 buffer was used for the list of hosts. I made this 1024 bytes long,
3941 thinking that nobody would ever have a list of hosts that long. Wrong.
3942 Somebody had a whole pile of complicated expansion conditions, and the
3943 string was silently truncated, leading to an expansion error. It turns out
3944 that it is easier to change to an unlimited length (owing to other changes
3945 that have happened since this code was originally written) than to build
3946 structure for giving a limitation error. The length of the item that
3947 expands into the list of hosts is now unlimited.
3949 47. The lsearch lookup could not handle data where the length of text line was
3950 more than 4095 characters. Such lines were truncated, leading to shortened
3951 data being returned. It should now handle lines of any length.
3953 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3954 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3957 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3958 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3959 no longer happens in comment lines. A list of replacements is now placed
3960 at the head of all of the source files, except those whose only change is
3961 to replace PERL_COMMAND in the very first #! line.
3963 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3964 messages on the queue, with a bottom-up merge sort, using code contributed
3965 by Michael Haardt. This should make operations like -bp somewhat faster on
3966 large queues. It won't affect queue runners, except when queue_run_in_order
3969 51. Installed eximstats 1.31 in the distribution.
3971 52. Added support for SRV lookups to the dnslookup router.
3973 53. If an ACL referred to $message_body or $message_body_end, the value was not
3974 reset for any messages that followed in the same SMTP session.
3976 54. The store-handling optimization for building very long strings was not
3977 differentiating between the different store pools. I don't think this
3978 actually made any difference in practice, but I've tidied it.
3980 55. While running the routers to verify a sender address, $sender_address
3981 was still set to the sender address. This is wrong, because when routing to
3982 send a bounce to the sender, it would be empty. Therefore, I have changed
3983 it so that, while verifying a sender address, $sender_address is set to <>.
3984 (There is no change to what happens when verifying a recipient address.)
3986 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3987 target A or AAAA records (if not already returned) without resetting the
3988 qualify_single or search_parents options of the DNS resolver. These are
3989 inappropriate in this case because the targets of MX and SRV records must
3990 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3991 target that, when qualified, matched something in the local domain. These
3992 two options are now turned off when doing these lookups.
3994 57. It seems that at least some releases of Reiserfs (which does not have the
3995 concept of a fixed number of inodes) returns zero and not -1 for the
3996 number of available inodes. This interacted badly with check_spool_inodes,
3997 which assumed that -1 was the "no such thing" setting. What I have done is
3998 to check that the total number of inodes is greater than zero before doing
3999 the test of how many are available.
4001 58. When a "warn" ACL statement has a log_message modifier, the message is
4002 remembered, and not repeated. This is to avoid a lot of repetition when a
4003 message has many recipients that cause the same warning to be written.
4004 Howewer, Exim was preserving the list of already written lines for an
4005 entire SMTP session, which doesn't seem right. The memory is now reset if a
4006 new message is started.
4008 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4009 debugging output unless log_rewrite was also set.
4011 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4012 of (int)(handle) when we know that handle contains (void *)(-1).
4014 61. The Exim daemon panic-logs an error return when it closes the incoming
4015 connection. However "connection reset by peer" seems to be common, and
4016 isn't really an error worthy of noting specially, so that particular error
4019 62. When Exim is trying to find all the local interfaces, it used to panic and
4020 die if the ioctl to get the interface flags failed. However, it seems that
4021 on at least one OS (Solaris 9) it is possible to have an interface that is
4022 included in the list of interfaces, but for which you get a failure error
4023 for this call. This happens when the interface is not "plumbed" into a
4024 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4025 failure of the "get flags" call assumes that the interface is down.
4027 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4028 makes life easier for people who are doing arithmetic on fields extracted
4029 from dates, where you often get leading zeros that should not be
4030 interpreted as octal.
4032 64. Added qualify_domain to the redirect router, to override the global
4035 65. If a pathologically long header line contained very many addresses (the
4036 report of this problem mentioned 10 000) and each of them was rewritten,
4037 Exim could use up a very large amount of memory. (It kept on making new
4038 copies of the header line as it rewrote, and never released the old ones.)
4039 At the expense of a bit more processing, the header rewriting function has
4040 been changed so that it no longer eats memory in this way.
4042 66. The generation of the Received: header has been moved from the time that a
4043 message starts to be received, to the time that it finishes. The timestamp
4044 in the Received: header should now be very close to that of the <= log
4045 line. There are two side-effects of this change:
4047 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4048 logged header lines no longer include the local Received: line, because
4049 it has not yet been created. The same applies to a copy of the message
4050 that is returned to a non-SMTP sender when a message is rejected.
4052 (b) When a filter file is tested using -bf, no additional Received: header
4053 is added to the test message. After some thought, I decided that this
4056 This change does not affect the value of $received_for. It is still set
4057 after address rewriting, but before local_scan() is called.
4059 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4061 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4062 gave an unhelpful panic error message, and a defer error. I have managed to
4063 change this behaviour so that it now rejects any supplied certificate,
4064 which seems right, as the list of acceptable certificates is empty.
4066 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4067 gave an unhelpful defer error. I have not managed to make this reject any
4068 supplied certificates, but the error message it gives is "no certificate
4069 supplied", which is not helpful.
4071 70. exigrep's output now also includes lines that are not associated with any
4072 message, but which match the given pattern. Implemented by a patch from
4073 Martin Sluka, which also tidied up the Perl a bit.
4075 71. Recipient callout verification, like sender verification, was using <> in
4076 the MAIL FROM command. This isn't really the right thing, since the actual
4077 sender may affect whether the remote host accepts the recipient or not. I
4078 have changed it to use the actual sender in the callout; this means that
4079 the cache record is now keyed on a recipient/sender pair, not just the
4080 recipient address. There doesn't seem to be a real danger of callout loops,
4081 since a callout by the remote host to check the sender would use <>.
4082 [SEE ABOVE: changed after hitting problems.]
4084 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4085 temporary errors. However, in the case of such a code being given after
4086 the end of a data transmission (i.e. after ".") Exim was failing to write
4087 a retry record for the message. (Yes, there was some broken host that was
4088 actually sending 8xx at this point.)
4090 73. An unknown lookup type in a host list could cause Exim to panic-die when
4091 the list was checked. (An example that provoked this was putting <; in the
4092 middle of a list instead of at the start.) If this happened during a DATA
4093 ACL check, a -D file could be left lying around. This kind of configuration
4094 error no longer causes Exim to die; instead it causes a defer errror. The
4095 incident is still logged to the main and panic logs.
4097 74. Buglet left over from Exim 3 conversion. The message "too many messages
4098 in one connection" was written to the rejectlog but not the mainlog, except
4099 when address rewriting (yes!) was being logged.
4101 75. Added write_rejectlog option.
4103 76. When a system filter was run not as root (that is, when system_filter_user
4104 was set), the values of the $n variables were not being returned to the
4105 main process; thus, they were not subsequently available in the $sn
4108 77. Added +return_path_on_delivery log selector.
4110 78. A connection timeout was being treated differently from recipients deferred
4111 when testing hosts_max_try with a message that was older than the host's
4112 retry timeout. (The host should not be counted, thus allowing all hosts to
4113 be tried at least once before bouncing.) This may have been the cause of an
4114 occasionally reported bug whereby a message would remain on the queue
4115 longer than the retry timeout, but would be bounced if a delivery was
4116 forced. I say "may" because I never totally pinned down the problem;
4117 setting up timeout/retry tests is difficult. See also the next item.
4119 79. The ultimate address timeout was not being applied to errors that involved
4120 a combination of host plus message (for example, a timeout on a MAIL
4121 command). When an address resolved to a number of possible hosts, and they
4122 were not all tried for each delivery (e.g. because of hosts_max_try), a
4123 message could remain on the queue longer than the retry timeout.
4125 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4128 81. Fixed an obscure SMTP outgoing bug which required at least the following
4129 conditions: (a) there was another message waiting for the same server;
4130 (b) the server returned 5xx to all RCPT commands in the first message so
4131 that the message was not completed; (c) the server dropped the connection
4132 or gave a negative response to the RSET that Exim sends to abort the
4133 transaction. The observed case was a dropped connection after DATA that had
4134 been sent in pipelining mode. That is, the server had advertised PIPELINING
4135 but was not implementing it correctly. The effect of the bug was incorrect
4136 behaviour, such as trying another host, and this could lead to a crash.
4142 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4143 and daemon.c were passed as pointers to ints; they should have been
4144 pointers to socklen_t variables (which are typically unsigned ints).
4146 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4149 3. Fixed a really odd bug that affected only the testing scheme; patching a
4150 certain fixed string in the binary changed the value of another string that
4151 happened to be identical to the end of the original first string.
4153 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4154 name", it returns that address as the IP address. On some operating
4155 systems (e.g. Solaris), it also passes back the IP address string as the
4156 "host name". However, on others (e.g. Linux), it passes back an empty
4157 string. Exim wasn't checking for this, and was changing the host name to an
4158 empty string, assuming it had been canonicized.
4160 5. Although rare, it is permitted to have more than one PTR record for a given
4161 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4162 all the names associated with an address, because they do in Solaris.
4163 However, it seems that they do not in Linux for data that comes from the
4164 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4165 I found this out when I moved to a new Linux workstation and tried to run
4166 the Exim test suite.
4168 To get round this problem I have changed the code so that it now does its
4169 own call to the DNS to look up PTR records when searching for a host name.
4170 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4171 addresses that are only in /etc/hosts are still found.
4173 This behaviour is, however, controlled by an option called host_lookup_
4174 order, which defaults to "bydns:byaddr". If people want to use the other
4175 order, or indeed, just use one or the other means of lookup, they can
4176 specify it in this variable.
4178 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4179 some operating systems, this comes back from gethostbyaddr() as an empty
4180 string, and this is what Exim used to test for. However, it seems that in
4181 other systems, "." is yielded. Exim now tests for this case too.
4183 7. The values of check_spool_space and check_log_space are now held internally
4184 as a number of kilobytes instead of an absolute number of bytes. If a
4185 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4186 kilobyte. This means that much larger values can be stored.
4188 8. Exim monitor: an attempt to get the action menu when not actually pointing
4189 at a message produces an empty menu entitled "No message selected". This
4190 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4191 no entries in it ("Shell widget menu has zero width and/or height"). So I
4192 have added a single, blank menu entry in this case.
4194 9. Added ${quote_local_part.
4196 10. MIME decoding is now applied to the contents of Subject: header lines when
4199 11. Now that a reference to $sender_host_address automatically causes a reverse
4200 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4201 host lookup before query-style lookups in lists that might use this
4202 variable. This has therefore been abolished, and the "net-" prefix is no
4203 longer necessary for query-style lookups.
4205 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4206 have been a typo for LFLAGS, so it has been changed.
4208 13. The install script calls Exim with "-C /dev/null" in order to find the
4209 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4210 to be output. Howeve, since Exim outputs its version number before the
4211 error, it didn't break the script. It just looked ugly. I fixed this by
4212 always allowing "-C /dev/null" if the caller is root.
4214 14. Ignore overlarge ACL variable number when reading spool file - insurance
4215 against a later release with more variables having written the file.
4217 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4218 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4219 this, as well as the form without the "IPv6" on the front.
4221 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4222 OS/Makefile-Darwin file.
4224 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4225 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4227 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4228 variable. However, some OS use ints for inodes. I've added cast to long int
4229 to get rid of the compiler warning.
4231 19. I had forgotten to lock out "/../" in configuration file names when
4232 ALT_CONFIG_PREFIX was set.
4234 20. Routers used for verification do not need to specify transports. However,
4235 if such a router generated a host list, and callout was configured, Exim
4236 crashed, because it could not find a port number from the (non-existent)
4237 transport. It now assumes port 25 in this circumstance.
4239 21. Added the -t option to exigrep.
4241 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4242 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4243 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4244 them is set, LOOKUP_LSEARCH is forced.
4246 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4248 24. Added sender and host information to the "rejected by local_scan()" log
4249 line; previously there was no indication of these.
4251 25. Added .include_if_exists.
4253 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4254 Linux. It turns out that not all file systems support this. Apparently some
4255 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4256 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4257 not supported on the file descriptor, is now ignored when Exim is trying to
4258 sync a directory. This applies only to Linux.
4260 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4262 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4263 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4266 29. Added support for maildirsize files from supplied patch (modified a bit).
4268 30. The use of :fail: followed by an empty string could lead Exim to respond to
4269 sender verification failures with (e.g.):
4271 550 Verification failed for <xxx>
4272 550 Sender verify failed
4274 where the first response line was missing the '-' that indicates it is not
4275 the final line of the response.
4277 31. The loop for finding the name of the user that called Exim had a hardwired
4278 limit of 10; it now uses the value of finduser_retries, which is used for
4279 all other user lookups.
4281 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4284 33. Exim was neglecting to zero errno before one call of strtol() when
4285 expanding a string and expecting an integer value. On some systems this
4286 resulted in spurious "integer overflow" errors. Also, it was casting the
4287 result into an int without checking.
4289 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4290 did not work. The code looks as if it has *never* worked, though it appears
4291 to have been documented since at least releast 1.62. I have made it work.
4293 35. The "timeout_DNS" error in retry rules, also documented since at least
4294 1.62, also never worked. As it isn't clear exactly what this means, and
4295 clearly it isn't a major issue, I have abolished the feature by treating it
4296 as "timeout", and writing a warning to the main and panic logs.
4298 36. The display of retry rules for -brt wasn't always showing the error code
4301 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4302 timeout_connect_A, timeout_connect_MX.
4304 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4305 to the empty sender.
4307 39. The daemon was not analysing the content of -oX till after it had closed
4308 stderr and disconnected from the controlling terminal. This meant that any
4309 syntax errors were only noted on the panic log, and the return code from
4310 the command was 0. By re-arranging the code a little, I've made the
4311 decoding happen first, so such errors now appear on stderr, and the return
4312 code is 1. However, the actual setting up of the sockets still happens in
4313 the disconnected process, so errors there are still only recorded on the
4316 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4317 connections (as happens on some IP stacks) was logged at start up time as
4318 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4319 it from "IPv6 and IPv4", which means that two separate sockets are being
4322 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4323 says whether AF_INET or AF_INET6 was passed as an argument.
4325 42. Exiwhat output was messed up when time zones were included in log
4328 43. Exiwhat now gives more information about the daemon's listening ports,
4329 and whether -tls-on-connect was used.
4331 44. The "port" option of the smtp transport is now expanded.
4333 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4334 silently ignored. Now an error message is written to the main and panic
4337 46. There's a new ACL modifier called "logwrite" which writes to a log file
4338 as soon as it is encountered.
4340 47. Added $local_user_uid and $local_user_gid at routing time.
4342 48. Exim crashed when trying to verify a sender address that was being
4345 49. Exim was recognizing only a space character after ".include". It now also
4346 recognizes a tab character.
4348 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4349 extracting the relevant information from the specification. The man page no
4350 longer contains scrambled data for the -d option, and I've added a section
4351 at the front about calling Exim under different names.
4353 51. Added "extra_headers" argument to the "mail" command in filter files.
4355 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4358 53. Installed eximstats 1.29.
4360 54. Added transport_filter_timeout as a generic transport option.
4362 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4363 Cc: header lines. This was required by RFC 822, but it not required by RFC
4366 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4367 incoming messages that did not have them. Now it does so only if the
4368 message originates locally, that is, if there is no associated remote host
4369 address. When Resent- header lines are present, this applies to the Resent-
4370 lines rather than the non-Resent- lines.
4372 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4373 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4375 58. Messages for configuration errors now include the name of the main
4376 configuration file - useful now that there may be more than one file in a
4377 list (.included file names were always shown).
4379 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4380 for those rare installations that do not start the daemon as root or run it
4381 setuid root. I've cut out the call to initgroups() if the daemon is not
4384 60. The Exim user and group can now be bound into the binary as text strings
4385 that are looked up at the start of Exim's processing.
4387 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4389 62. Added $mailstore_basename variable.
4391 63. Installed patch to sieve.c from Michael Haardt.
4393 64. When Exim failed to open the panic log after failing to open the main log,
4394 the original message it was trying to log was written to stderr and debug
4395 output, but if they were not available (the usual case in production), it
4396 was lost. Now it is written to syslog before the two lines that record the
4397 failures to open the logs.
4399 65. Users' Exim filters run in subprocesses under the user's uid. It is
4400 possible for a "deliver" command or an alias in a "personal" command to
4401 provoke an address rewrite. If logging of address rewriting is configured,
4402 this fails because the process is not running as root or exim. There may be
4403 a better way of dealing with this, but for the moment (because 4.30 needs
4404 to be released), I have disabled address rewrite logging when running a
4405 filter in a non-root, non-exim process.
4411 1. The buildconfig auxiliary program wasn't quoting the value set for
4412 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4413 not defined. This bug was masked in 4.22 by the effect that was fixed in
4416 2. Some messages that were rejected after a message id was allocated were
4417 shown as "incomplete" by exigrep. It no longer does this for messages that
4418 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4420 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4421 have allow_domain_literals set, the ID did not get logged in the <= line.
4422 Domain literals are now always recognized in Message-ID: header lines.
4424 4. The first argument for a ${extract expansion item is the key name or field
4425 number. Leading and trailing spaces in this item were not being ignored,
4426 causing some misleading effects.
4428 5. When deliver_drop_privilege was set, single queue runner processes started
4429 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4430 same command in the process it spins off) were not dropping privilege.
4432 6. When the daemon running as "exim" started a queue runner, it always
4433 re-executed Exim in the spun-off process. This is a waste of effort when
4434 deliver_drop_privilege is set. The new process now just calls the
4435 queue-runner function directly.
4441 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4444 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4445 ignored. Though the use of -oP was forcing the writing of a pid file, it
4446 was always written to the default place.
4448 3. If the message "no IP address found for host xxxx" is generated during
4449 incoming verification, it is now followed by identification of the incoming
4450 connection (so you can more easily find what provoked it).
4452 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4454 5. Added some features to "harden" Exim a bit more against certain attacks:
4456 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4457 be put in Local/Makefile. This is like the never_users runtime option,
4458 but it cannot be overridden. The default setting is "root".
4460 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4461 prefix string with which any file named in a -C command line option
4464 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4465 is retained for -C and -D only if the caller of Exim is root. Without
4466 it, the exim user may also use -C and -D and retain privilege.
4468 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4469 command line option is disabled.
4471 6. Macro names set by the -D option must start with an upper case letter, just
4472 like macro names defined in the configuration file.
4474 7. Added "dereference=" facility to LDAP.
4476 8. Two instances of the typo "uknown" in the source files are fixed.
4478 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4479 the Configure-Makefile script screwed up while processing it.
4481 10. Incorporated PCRE 4.4.
4483 11. The SMTP synchronization check was not operating right at the start of an
4484 SMTP session. For example, it could not catch a HELO sent before the client
4485 waited for the greeting. There is now a check for outstanding input at the
4486 point when the greeting is written. Because of the duplex, asynchronous
4487 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4488 way, but not yet received, when the check is performed.
4490 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4491 on TCP/IP sockets, because this apparently causes some broken clients to
4494 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4495 unchanged) from the Cygwin maintainer.
4497 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4498 was supported instead of testing for mbx. Effectively a typo.
4500 15. The spa authenticator server code was not checking that the input it
4501 received was valid base64.
4503 16. The debug output line for the "set" modifier in ACLs was not showing the
4504 name of the variable that was being set.
4506 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4508 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4509 lookup on its own. Something else was needed to trigger the lookup. For
4510 example, a match in host_lookup or the need for a host name in a host list.
4511 Now, if $sender_host_name is referenced and the host name has not yet been
4512 looked up, a lookup is performed. If the lookup fails, the variable remains
4513 empty, and $host_lookup_failed is set to "1".
4515 19. Added "eqi" as a case-independent comparison operator.
4517 20. The saslauthd authentication condition could segfault if neither service
4518 nor realm was specified.
4520 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4521 error message that was logged was misleading, and incoming SMTP
4522 connections were dropped. The message is now more accurate, and temporary
4523 errors are given to SMTP connections.
4525 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4526 (except RSET) with a 421 error, until QUIT is received. However, it was
4527 failing to send a response to QUIT.
4529 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4530 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4531 ACL is now run after the helo_try_verify_hosts code.
4533 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4534 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4535 case-independent, so other case variants are also recognized). Apparently
4536 some systems use these upper case variants.
4538 25. If more than two messages were waiting for the same host, and a transport
4539 filter was specified for the transport, Exim sent two messages over the
4540 same TCP/IP connection, and then failed with "socket operation on non-
4541 socket" when it tried to send the third.
4543 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4545 27. The extern definition of crypt16() in expand.c was not being excluded when
4546 the OS had its own crypt16() function.
4548 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4549 as a preferred synonym for return_size_limit, both as an option and as an
4552 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4554 30. Changed the default configuration ACL to relax the local part checking rule
4555 for addresses that are not in any local domains. For these addresses,
4556 slashes and pipe symbols are allowed within local parts, but the sequence
4557 /../ is explicitly forbidden.
4559 31. SPA server authentication was not clearing the challenge buffer before
4562 32. log_message in a "warn" ACL statement was writing to the reject log as
4563 well as to the main log, which contradicts the documentation and doesn't
4564 seem right (because no rejection is happening). So I have stopped it.
4566 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4567 However, I am unable to do any testing of this.
4569 34. Fixed an infelicity in the appendfile transport. When checking directories
4570 for a mailbox, to see if any needed to be created, it was accidentally
4571 using path names with one or more superfluous leading slashes; tracing
4572 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4574 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4575 contents are added to the log line that is written for every discarded
4576 recipient. (Previously a log_message setting was ignored.)
4578 36. The ${quote: operator now quotes the string if it is empty.
4580 37. The install script runs exim in order to find its version number. If for
4581 some reason other than non-existence or emptiness, which it checks, it
4582 could not run './exim', it was installing it with an empty version number,
4583 i.e. as "exim-". This error state is now caught, and the installation is
4586 38. An argument was missing from the function that creates an error message
4587 when Exim fails to connect to the socket for saslauthd authentication.
4588 This could cause Exim to crash, or give a corrupted message.
4590 39. Added isip, isip4, and isip6 to ${if conditions.
4592 40. The ACL variables $acl_xx are now saved with the message, and can be
4593 accessed later in routers, transports, and filters.
4595 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4596 strings in the file are not string-expanded.
4598 42. If a MAIL command specified a SIZE value that was too large to fit into an
4599 int variable, the check against message_size_limit failed. Such values are
4600 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4601 day this will have to be increased, but I don't think I want to be around
4602 when emails are that large.
4609 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4610 iconv() is not standard in FreeBSD.
4612 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4613 IPv6 enabled. The observed symptom was a segmentation fault on return from
4614 the function os_common_find_running_interfaces() in src/os.c.
4616 3. In the check_special_case() function in daemon.c I had used "errno" as an
4617 argument name, which causes warnings on some systems. This was basically a
4618 typo, since it was named "eno" in the comments!
4620 4. The code that waits for the clock to tick (at a resolution of some fraction
4621 of a second) so as to ensure message-id uniqueness was always waiting for
4622 at least one whole tick, when it could have waited for less. [This is
4623 almost certainly not relevant at current processor speeds, where it is
4624 unlikely to ever wait at all. But we try to future-proof.]
4626 5. The function that sleeps for a time interval that includes fractions of a
4627 second contained a race. It did not block SIGALRM between setting the
4628 timer, and suspending (a couple of lines later). If the interval was short
4629 and the sigsuspend() was delayed until after it had expired, the suspension
4630 never ended. On busy systems this could lead to processes getting stuck for
4633 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4634 process, before it forks any delivery processes. The open lookup caching
4635 mechanism meant that the open file or database connection was passed into
4636 the delivery process. The problem was that delivery processes always tidy
4637 up cached lookup data. This could cause a problem for the next delivery
4638 process started by the queue runner, because the external queue runner
4639 process does not know about the closure. So the next delivery process
4640 still has data in the lookup cache. In the case of a file lookup, there was
4641 no problem because closing a file descriptor in a subprocess doesn't affect
4642 the parent. However, if the lookup was caching a connection to a database,
4643 the connection was closed, and the second delivery process was likely to
4644 see errors such as "PGSQL: query failed: server closed the connection
4645 unexpectedly". The problem has been fixed by closing all cached lookups
4646 in a queue runner before running a delivery process.
4648 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4649 seem to have the "const" qualifier which it has on other OS. I've
4652 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4653 *of the same type* (client or server) with the same public name that an
4654 error should be diagnosed.
4656 9. When Exim looked up a host name for an IP address, but failed to find the
4657 original IP address when looking up the host name (a safety check), it
4658 output the message "<ip address> does not match any IP for NULL", which was
4659 confusing, to say the least. The bug was that the host name should have
4660 appeared instead of "NULL".
4662 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4663 user that is built into the binary, and the -C or -D options is used, root
4664 privilege is dropped before the configuration file is read. In addition,
4665 logging is switched to stderr instead of the normal log files. If the
4666 configuration then re-defines the Exim user, the unprivileged environment
4667 is probably not what is expected, so Exim logs a panic warning message (but
4670 However, if deliver_drop_privilege is set, the unprivileged state may well
4671 be exactly what is intended, so the warning has been cut out in that case,
4672 and Exim is allowed to try to write to its normal log files.
4678 1. smtp_return_error_details was not giving details for temporary sender
4679 or receiver verification errors.
4681 2. Diagnose a configuration error if two authenticators have the same public
4684 3. Exim used not to create the message log file for a message until the first
4685 delivery attempt. This could be confusing when incoming messages were held
4686 for policy or load reasons. The message log file is now created at the time
4687 the message is received, and an initial "Received" line is written to it.
4689 4. The automatically generated man page for command line options had a minor
4690 bug that caused no ill effects; however, a more serious problem was that
4691 the procedure for building the man page automatically didn't always
4692 operate. Consequently, release 4.20 contains an out-of-date version. This
4693 shouldn't happen again.
4695 5. When building Exim with embedded Perl support, the script that builds the
4696 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4697 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4699 6. The freeze_tell option was not being used for messages that were frozen on
4700 arrival, either by an ACL or by local_scan().
4702 7. Added the smtp_incomplete_transaction log selector.
4704 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4705 was accepting AUTH without a new EHLO.
4707 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4708 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4711 10. Logging of TCP/IP connections (when configured) now happens in the main
4712 daemon process instead of the child process, so that the TCP/IP connection
4713 count is more accurate (but it can never be perfect).
4715 11. The use of "drop" in a nested ACL was not being handled correctly in the
4716 outer ACL. Now, if condition failure induced by the nested "drop" causes
4717 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4718 or "require"), the connection is dropped.
4720 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4721 that yield "discard" can now be used with an "accept" or a "discard" verb,
4722 but an error is generated for any others (because I can't see a useful way
4723 to define what should happen).
4725 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4726 are now processed in the same way as lines in the Exim configuration file.
4727 In particular, continuation lines are supported.
4729 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4731 15. Added -ti meaning -t -i.
4733 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4734 domains, and warn by logging if others are found.
4736 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4737 returned from GIFCONF when Exim is trying to find the list of interfaces on
4738 a host. The code in os.c has been modified to copy each ifreq to an aligned
4739 structure in all cases.
4741 Also, in some cases, the returned ifreq's were being copied to a 'struct
4742 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4743 means the last couple of bytes of an IPv6 address could be chopped if the
4744 ifreq contained only a normal sockaddr (14 bytes storage).
4746 18. Named domain lists were not supported in the hosts_treat_as_local option.
4747 An entry such as +xxxx was not recognized, and was treated as a literal
4750 19. Ensure that header lines added by a DATA ACL are included in the reject log
4751 if the ACL subsequently rejects the message.
4753 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4754 MD5 (which is deprecated).
4756 21. When testing a filter file using -bf, Exim was writing a message when it
4757 took the sender from a "From " line in the message, but it was not doing so
4758 when it took $return_path from a Return-Path: header line. It now does.
4760 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4761 with a valid header line field name (a series of printing characters
4762 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4764 23. Changed "disc" in the source to "disk" to conform to the documentation and
4765 the book and for uniformity.
4767 24. Ignore Sendmail's -Ooption=value command line item.
4769 25. When execve() failed while trying to run a command in a pipe transport,
4770 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4771 could be confused with a return value of 69 from the command itself. This
4772 has been changed to 127, the value the shell returns if it is asked to run
4773 a non-existent command. The wording for the related log line suggests a
4774 non-existent command as the problem.
4776 26. If received_header_text expands to an empty string, do not add a Received:
4777 header line to the message. (Well, it adds a token one on the spool, but
4778 marks it "old" so that it doesn't get used or transmitted.)
4780 27. Installed eximstats 1.28 (addition of -nt option).
4782 28. There was no check for failure on the call to getsockname() in the daemon
4783 code. This can fail if there is a shortage of resources on the system, with
4784 ENOMEM, for example. A temporary error is now given on failure.
4786 29. Contrary to the C standard, it seems that in some environments, the
4787 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4788 program. Exim now does this explicitly; it affects the formatting of
4789 timestamps using strftime().
4791 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4792 complaints. I've now initialized all the variables, to avoid this.
4794 32. Header lines added by a system filter were not being "seen" during
4795 transport-time rewrites.
4797 33. The info_callback() function passed to OpenSSL is set up with type void
4798 (*)(SSL *, int, int), as described somewhere. However, when calling the
4799 function (actually a macro) that sets it up, the type void(*)() is
4800 expected. I've put in a cast to prevent warnings from picky compilers.
4802 34. If a DNS black list lookup found a CNAME record, but there were no A
4803 records associated with the domain it pointed at, Exim crashed.
4805 35. If a DNS black list lookup returned more than one A record, Exim ignored
4806 all but the first. It now scans all returned addresses if a particular IP
4807 value is being sought. In this situation, the contents of the
4808 $dnslist_value variable are a list of all the addresses, separated by a
4811 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4812 to accept a host name and all its aliases if the forward lookup for any of
4813 them yielded the IP address of the incoming connection. Now it accepts only
4814 those names whose forward lookup yields the correct IP address. Any other
4815 names are discarded. This closes a loophole whereby a rogue DNS
4816 administrator could create reverse DNS records to break through a
4817 wildcarded host restriction in an ACL.
4819 37. If a user filter or a system filter that ran in a subprocess used any of
4820 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4821 the wrong values were passed to the pipe command ($thisaddress had the
4822 value of $0, $0 had the value of $1, etc). This bug was introduced by
4823 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4825 38. Improved the line breaking for long SMTP error messages from ACLs.
4826 Previously, if there was no break point between 40 and 75 characters, Exim
4827 left the rest of the message alone. Two changes have been made: (a) I've
4828 reduced the minimum length to 35 characters; (b) if it can't find a break
4829 point between 35 and 75 characters, it looks ahead and uses the first one
4830 that it finds. This may give the occasional overlong line, but at least the
4831 remaining text gets split now.
4833 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4834 file descriptors might be low, and that setting 1000 would always raise it.
4835 It turns out that in some environments, the limit is already over 1000 and
4836 that lowering it causes trouble. So now Exim takes care not to decrease it.
4838 40. When delivering a message, the value of $return_path is set to $sender_
4839 address at the start of routing (routers may change the value). By an
4840 oversight, this default was not being set up when an address was tested by
4841 -bt or -bv, which affected the outcome if any router or filter referred to
4844 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4845 log only when "message" or "log_message" are set. However, if one of the
4846 conditions was an address verification, or a call to a nested ACL, the
4847 messages generated by the underlying test were being passed through. This
4848 no longer happens. The underlying message is available in $acl_verify_
4849 message for both "message" and "log_message" expansions, so it can be
4850 passed through if needed.
4852 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4853 new expansion $bh_ to give the encoded byte string without charset
4854 translation. Translation happens only if iconv() is available; HAVE_ICONV
4855 indicates this at build time. HEADERS_CHARSET gives the charset to
4856 translate to; headers_charset can change it in the configuration, and
4857 "headers charset" can change it in an individual filter file.
4859 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4860 that creates RFC 2047 encoded "words" labels them as that charset instead
4861 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4862 expansion operator; (ii) when Exim creates a From: line for a local
4863 message; (iii) when a header line is rewritten to include a "phrase" part.
4865 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4866 buggy, causing it to skip the first lines of messages whose message ID
4867 ended in 'D'. This would not have bitten before Exim release 4.14, because
4868 message IDs were unlikely to end in 'D' before then. The effect was to have
4869 incorrect size information for certain domains.
4871 45. #include "config.h" was missing at the start of the crypt16.c module. This
4872 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4875 46. If there was a timeout during a "random" callout check, Exim treated it as
4876 a failure of the random address, and carried on sending RSET and the real
4877 address. If the delay was just some slowness somewhere, the response to the
4878 original RCPT would be taken as a response to RSET and so on, causing
4879 mayhem of various kinds.
4881 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4882 when I implemented it. It didn't allow for the fact that some option values
4883 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4884 the right test for positive values.
4886 48. Domain names in DNS records are case-independent. Exim always looks them up
4887 in lower case. Some resolvers return domain names in exactly the case they
4888 appear in the zone file, that is, they may contain uppercase letters. Not
4889 all resolvers do this - some return always lower case. Exim was treating a
4890 change of case by a resolver as a change of domain, similar to a widening
4891 of a domain abbreviation. This triggered its re-routing code and so it was
4892 trying to route what was effectively the same domain again. This normally
4893 caused routing to fail (because the router wouldn't handle the domain
4894 twice). Now Exim checks for this case specially, and just changes the
4895 casing of the domain that it ultimately uses when it transmits the message
4898 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4901 50. If a filter generated a file delivery with a non-absolute name (possible if
4902 no home directory exists for the router), the forbid_file option was not
4905 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4906 the existing equality matching.
4908 52. Exim was using ints instead of ino_t variables in some places where it was
4909 dealing with inode numbers.
4911 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4912 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4913 called TMPDIR, and if it finds it is different, it changes its value.
4915 54. The smtp_printf() function is now made available to local_scan() so
4916 additional output lines can be written before returning. There is also an
4917 smtp_fflush() function to enable the detection of a dropped connection.
4918 The variables smtp_input and smtp_batched_input are exported to
4921 55. Changed the default runtime configuration: the message "Unknown user"
4922 has been removed from the ACL, and instead placed on the localuser router,
4923 using the cannot_route_message feature. This means that any verification
4924 failures that generate their own messages won't get overridden. Similarly,
4925 the "Unrouteable address" message that was in the ACL for unverifiable
4926 relay addresses has also been removed.
4928 56. Added hosts_avoid_esmtp to the smtp transport.
4930 57. The exicyclog script was not checking for the esoteric option
4931 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4932 will work only if exicyclog is run under the appropriate euid.
4934 58. Following a discussion on the list, the rules by which Exim recognises line
4935 endings on incoming messages have been changed. The -dropcr and drop_cr
4936 options are now no-ops, retained only for backwards compatibility. The
4937 following line terminators are recognized: LF CRLF CR. However, special
4938 processing applies to CR:
4940 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4941 nor a local message in the state where . is a terminator.
4943 (ii) If a bare CR is encountered in a header line, an extra space is added
4944 after the line terminator so as not to end the header. The reasoning
4945 behind this is that bare CRs in header lines are most likely either
4946 to be mistakes, or people trying to play silly games.
4948 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4949 was being incorrectly given as 18 bytes larger than it should have been.
4950 This is a VOB (very old bug).
4952 60. This may never have affected anything current, but just in case it has:
4953 When the local host is found other than at the start of a list of hosts,
4954 the local host, those with the same MX, and any that follow, are discarded.
4955 When the list in question was part of a longer list of hosts, the following
4956 hosts (not currently being processed) were also being discarded. This no
4957 longer happens. I'm not sure if this situation could ever has previously
4960 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4963 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4964 header. This is something that is recommended in a new Internet Draft, and
4965 is something that is documented as being done by Sendmail. There are two
4966 possible values. For messages generated by the autoreply transport, Exim
4969 Auto-Submitted: auto-replied
4971 whereas for all other generated messages (e.g. bounces) it adds
4973 Auto-Submitted: auto-generated
4975 63. The "personal" condition in filters now includes a test for the
4976 Auto-Submitted: header. If it contains the string "auto-" the message it
4977 not considered personal.
4979 64. Added rcpt_include_affixes as a generic transport option.
4981 65. Added queue_only_override (default true).
4983 66. Added the syslog_duplication option.
4985 67. If what should have been the first header line of a message consisted of
4986 a space followed by a colon, Exim was mis-interpreting it as a header line.
4987 It isn't of course - it is syntactically invalid and should therefore be
4988 treated as the start of the message body. The misbehaviour could have
4989 caused a number of strange effects, including loss of data in subsequent
4990 header lines, and spool format errors.
4992 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4993 client host had authenticated. This control can now be exercised by an ACL
4994 for more flexibility.
4996 69. By default, callouts do not happen when testing with -bh. There is now a
4997 variant, -bhc, which does actually run the callout code, including
4998 consulting and updating the callout cache.
5000 70. Added support for saslauthd authentication, courtesy of Alexander
5003 71. If statvfs() failed on the spool or log directories while checking their
5004 size for availability, Exim confusingly gave the error "space shortage".
5005 Furthermore, in debugging mode it crashed with a floating point exception.
5006 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5007 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5008 serious problem, Exim now writes to the main and panic logs when this
5009 happens, with details of the failure. It then refuses to accept the
5010 incoming message, giving the message "spool directory problem" or "log
5011 directory problem" with a 421 code for SMTP messages.
5013 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5014 0, 1, and 2 exist, because some OS complain for execs without them (see
5015 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5016 descriptors. However, the code omitted to check that the open succeeded,
5017 causing mysterious errors if for some reason the permissions on /dev/null
5018 got screwed. Now Exim writes a message to the main and panic logs, and
5019 bombs out if it can't open /dev/null.
5021 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5022 interact so that it is all more flexible. It is supposed to remain
5023 backwards compatible. Also added extra_local_interfaces.
5025 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5026 to bomb out with an assertion failure - to the client this appears as a
5027 connection drop. This problem occurs in the part of the code that was taken
5028 from the Samba project. Fortunately, the assertion is in a very simple
5029 function, so I have fixed this by reproducing the function inline in the
5030 one place where it is called, and arranging for authentication to fail
5031 instead of killing the process with assert().
5033 75. The SPA client code was not working when the server requested OEM rather
5034 than Unicode encoding.
5036 76. Added code to make require_files with a specific uid setting more usable in
5037 the case where statting the file as root fails - usually a non-root-mounted
5038 NFS file system. When this happens and the failure is EACCES, Exim now
5039 forks a subprocess and does the per-uid checking as the relevant uid.
5041 77. Added process_log_path.
5043 78. If log_file_path was not explicitly set, a setting of check_log_space or
5044 check_log_inodes was ignored.
5046 79. If a space check for the spool or log partitions fails, the incident is now
5047 logged. Of course, in the latter case the data may get lost...
5049 80. Added the %p formatting code to string_format() so that it can be used to
5050 print addresses in debug_print(). Adjusted all the address printing in the
5051 debugging in store.c to use %p rather than %d.
5053 81. There was a concern that a line of code in smtp_in.c could overflow a
5054 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5055 initially expressed, the concern was not well-founded, because trailing
5056 spaces are removed early. However, if the trailing spaces were followed by
5057 a NULL, they did not get removed, so the overflow was possible. Two fixes
5060 (a) I re-wrote the offending code in a cleaner fashion.
5061 (b) If an incoming SMTP command contains a NULL character, it is rejected
5064 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5065 runs initgroups(), so that if the Exim user is in any additional groups,
5066 they will be used during message reception.
5072 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob