1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.612 2010/05/30 18:01:48 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
42 NM/08 Bugzilla 967. DKIM DNS TXT record bug fix
45 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
46 MBX locking. Notification and patch from Dan Rosenberg
52 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body
54 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
56 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
58 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
60 NM/04 Bugzilla 915: Flag broken perl installation during build.
66 TK/01 Added patch by Johannes Berg that expands the main option
67 "spamd_address" if it starts with a dollar sign.
69 TK/02 Write list of recipients to X-Envelope-Sender header when building
70 the mbox-format spool file for content scanning (suggested by Jakob
73 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
74 (http://www.dcc-servers.net/) support via dccifd. Activated by
75 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
77 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
78 by Mark Daniel Reidel <mr@df.eu>.
80 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
81 When building exim an external PCRE library is now needed -
82 PCRE is a system library on the majority of modern systems.
83 See entry on PCRE_LIBS in EDITME file.
85 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
86 conversation. Added nologin parameter to request.
87 Patch contributed by Kirill Miazine
89 TF/01 Do not log submission mode rewrites if they do not change the address.
91 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
93 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
94 log files in place. Contributed by Roberto Lima
96 NM/04 Bugzilla 667: close socket used by dovecot authenticator
98 TF/03 Bugzilla 615: When checking the local_parts router precondition
99 after a local_part_suffix or local_part_prefix option, Exim now
100 does not use the address's named list lookup cache, since this
101 contains cached lookups for the whole local part.
103 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
104 Robert Millan. Documentation is in experimental-spec.txt
106 TF/04 Bugzilla 668: Fix parallel build (make -j).
108 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000
110 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
111 Patch provided by Jan Srzednicki
113 TF/05 Leading white space used to be stripped from $spam_report which
114 wrecked the formatting. Now it is preserved.
116 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
117 that they are available at delivery time.
119 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
121 TF/08 TLS error reporting now respects the incoming_interface and
122 incoming_port log selectors.
124 TF/09 Produce a more useful error message if an SMTP transport's hosts
125 setting expands to an empty string.
127 NM/06 Bugzilla 744: EXPN did not work under TLS.
128 Patch provided by Phil Pennock
130 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
131 Patch provided by Richard Godbee
133 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
134 acl_smtp_notquit, added index entry.
136 NM/09 Bugzilla 787: Potential buffer overflow in string_format
137 Patch provided by Eugene Bujak
139 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
140 Patch provided by Maxim Dounin
142 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
143 Patch provided by Phil Pennock
145 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
147 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
148 Patch provided by Brad "anomie" Jorsch
150 NM/14 Bugzilla 622: Added timeout setting to transport filter.
151 Patch provided by Dean Brooks
153 TK/05 Add native DKIM support (does not depend on external libraries).
155 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
156 Patch provided by Graeme Fowler
158 NM/16 Bugzilla 851: Documentation example syntax fix.
160 NM/17 Changed NOTICE file to remove references to embedded PCRE.
162 NM/18 Bugzilla 894: Fix issue with very long lines including comments in lsearch
164 NM/19 Bugzilla 745: TLS version reporting
165 Patch provided by Phil Pennock
167 NM/20 Bugzilla 167: bool: condition support
168 Patch provided by Phil Pennock
170 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken clients
171 Patch provided by Phil Pennock
173 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date
174 Patch provided by Brad "anomie" Jorsch
176 NM/23 Bugzilla 687: Fix misparses in eximstats
177 Patch provided by Heiko Schlittermann
179 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid
180 Patch provided by Heiko Schlittermann
182 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file
183 plus update to original patch.
185 NM/26 Bugzilla 799: Documentation correction for ratelimit
187 NM/27 Bugzilla 802: Improvements to local interface IP addr detection
188 Patch provided by David Brownlee
190 NM/28 Bugzilla 807: Improvements to LMTP delivery logging
192 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes
194 NM/30 Bugzilla 888: TLS documentation bugfixes
196 NM/31 Bugzilla 896: Dovecot buffer overrun fix
198 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
199 Unlike the original bugzilla I have changed all shell scripts in src tree
201 NM/33 Bugzilla 898: Transport filter timeout fix
202 Patch by Todd Rinaldo
204 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches
205 Patch by Serge Demonchaux
207 NM/35 Bugzilla 39: Base64 decode bug fixes
208 Patch by Jakob Hirsch
210 NM/36 Bugzilla 909: Correct connect() call in dcc code
212 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling
214 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed
216 NM/39 Bugzilla 911: Fixed MakeLinks build script
222 TK/01 Add preliminary DKIM support. Currently requires a forked version of
223 ALT-N's libdkim that I have put here:
224 http://duncanthrax.net/exim-experimental/
226 Note to Michael Haardt: I had to rename some vars in sieve.c. They
227 were called 'true' and it seems that C99 defines that as a reserved
228 keyword to be used with 'bool' variable types. That means you could
229 not include C99-style headers which use bools without triggering
230 build errors in sieve.c.
232 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
233 as mailq or other aliases. Changed the --help handling significantly
234 to do whats expected. exim_usage() emits usage/help information.
236 SC/01 Added the -bylocaldomain option to eximstats.
238 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
240 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
242 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
248 PH/01 Another patch from the Sieve maintainer.
250 PH/02 When an IPv6 address is converted to a string for single-key lookup
251 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
252 dots are used instead of colons so that keys in lsearch files need not
253 contain colons. This was done some time before quoting was made available
254 in lsearch files. However, iplsearch files do require colons in IPv6 keys
255 (notated using the quote facility) so as to distinguish them from IPv4
256 keys. This meant that lookups for IP addresses in host lists did not work
257 for iplsearch lookups.
259 This has been fixed by arranging for IPv6 addresses to be expressed with
260 colons if the lookup type is iplsearch. This is not incompatible, because
261 previously such lookups could never work.
263 The situation is now rather anomolous, since one *can* have colons in
264 ordinary lsearch keys. However, making the change in all cases is
265 incompatible and would probably break a number of configurations.
267 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
270 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
271 conversion specification without a maximum field width, thereby enabling
272 a rogue spamd server to cause a buffer overflow. While nobody in their
273 right mind would setup Exim to query an untrusted spamd server, an
274 attacker that gains access to a server running spamd could potentially
275 exploit this vulnerability to run arbitrary code as the Exim user.
277 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
278 $primary_hostname instead of what libspf2 thinks the hosts name is.
280 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
281 a directory entry by the name of the lookup key. Previously, if a
282 symlink pointed to a non-existing file or a file in a directory that
283 Exim lacked permissions to read, a lookup for a key matching that
284 symlink would fail. Now it is enough that a matching directory entry
285 exists, symlink or not. (Bugzilla 503.)
287 PH/03 The body_linecount and body_zerocount variables are now exported in the
290 PH/04 Added the $dnslist_matched variable.
292 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
293 This means they are set thereafter only if the connection becomes
296 PH/06 Added the client_condition to authenticators so that some can be skipped
297 by clients under certain conditions.
299 PH/07 The error message for a badly-placed control=no_multiline_responses left
300 "_responses" off the end of the name.
302 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
304 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
305 (without spaces) instead of just copying the configuration text.
307 PH/10 Added the /noupdate option to the ratelimit ACL condition.
309 PH/11 Added $max_received_linelength.
311 PH/12 Added +ignore_defer and +include_defer to host lists.
313 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
314 way in which PCRE > 7.0 is built.
316 PH/14 Implemented queue_only_load_latch.
318 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
319 MAIL command. The effect was to mangle the value on 64-bit systems.
321 PH/16 Another patch from the Sieve maintainer.
323 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
325 PH/18 If a system quota error occurred while trying to create the file for
326 a maildir delivery, the message "Mailbox is full" was not appended to the
327 bounce if the delivery eventually timed out. Change 4.67/27 below applied
328 only to a quota excession during the actual writing of the file.
330 PH/19 It seems that peer DN values may contain newlines (and other non-printing
331 characters?) which causes problems in log lines. The DN values are now
332 passed through string_printing() before being added to log lines.
334 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
335 and InterBase are left for another time.)
337 PH/21 Added message_body_newlines option.
339 PH/22 Guard against possible overflow in moan_check_errorcopy().
341 PH/23 POSIX allows open() to be a macro; guard against that.
343 PH/24 If the recipient of an error message contained an @ in the local part
344 (suitably quoted, of course), incorrect values were put in $domain and
345 $local_part during the evaluation of errors_copy.
351 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
352 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
355 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
356 issue a MAIL command.
358 PH/02 In an ACL statement such as
360 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
362 if a client was not listed at all, or was listed with a value other than
363 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
364 the condition was not true (as it should be), so access was not denied.
365 The bug was that the ! inversion was incorrectly passed on to the second
366 item. This has been fixed.
368 PH/03 Added additional dnslists conditions == and =& which are different from
369 = and & when the dns lookup returns more than one IP address.
371 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
372 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
374 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
375 FSYNC, which compiles an option called disable_fsync that allows for
376 bypassing fsync(). The documentation is heavily laced with warnings.
378 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
380 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
381 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
382 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
383 including adding "make clean"; (3) Added -fPIC when compiling the test
384 dynamically loaded module, to get rid of a warning.
386 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
387 message fails, move_frozen_messages = true and ignore_bounce_errors_after
388 = 0s. The bug is otherwise harmless.
390 PH/07 There was a bug in the dovecot authenticator such that the value of
391 $auth1 could be overwritten, and so not correctly preserved, after a
392 successful authentication. This usually meant that the value preserved by
393 the server_setid option was incorrect.
395 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
397 PH/09 Installed PCRE release 7.0.
399 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
400 run for batched SMTP input. It is now run at the start of every message
401 in the batch. While fixing this I discovered that the process information
402 (output by running exiwhat) was not always getting set for -bs and -bS
403 input. This is fixed, and it now also says "batched" for BSMTP.
405 PH/11 Added control=no_pipelining.
407 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
408 patch, slightly modified), and move the expansion of helo_data till after
409 the connection is made in the smtp transport (so it can use these
412 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
414 PH/14 Added log_selector = +pid.
416 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
418 PH/16 Add ${if forany and ${if forall.
420 PH/17 Added dsn_from option to vary the From: line in DSNs.
422 PH/18 Flush SMTP output before performing a callout, unless control =
423 no_callout_flush is set.
425 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
426 was true (the default) a successful delivery failed to delete the retry
427 item, thus causing premature timeout of the address. The bug is now
430 PH/20 Added hosts_avoid_pipelining to the smtp transport.
432 PH/21 Long custom messages for fakedefer and fakereject are now split up
433 into multiline reponses in the same way that messages for "deny" and
434 other ACL rejections are.
436 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
437 with slight modification.
439 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
440 draft, changing the syntax and factoring some duplicate code".
442 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
443 for deliveries of the second and subsequent messages over the same SMTP
446 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
447 ${reduce, with only minor "tidies".
449 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
451 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
452 expansion side effects.
454 PH/27 When a message times out after an over-quota error from an Exim-imposed
455 quota, the bounce message says "mailbox is full". This message was not
456 being given when it was a system quota that was exceeded. It now should
459 MH/03 Made $recipients available in local_scan(). local_scan() already has
460 better access to the recipient list through recipients_list[], but
461 $recipients can be useful in postmaster-provided expansion strings.
463 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
464 in the case of a MAIL command with additional options following the
465 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
466 were accidentally chopped off.
468 PH/29 SMTP synchronization checks are implemented when a command is read -
469 there is a check that no more input is waiting when there shouldn't be
470 any. However, for some commands, a delay in an ACL can mean that it is
471 some time before the response is written. In this time, more input might
472 arrive, invalidly. So now there are extra checks after an ACL has run for
473 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
474 pipelining has not been advertised.
476 PH/30 MH's patch to allow iscntrl() characters to be list separators.
478 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
479 returned in the SMTP response when smtp_return_error_details was false.
482 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
483 instead of the C I/O that was originally supplied, because problems were
486 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
487 Exim which did not show up earlier: it was assuming that a call to
488 SSL_CTX_set_info_callback() might give an error value. In fact, there is
489 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
490 was a macro that became an assignment, so it seemed to work. This has
491 changed to a proper function call with a void return, hence the compile
492 error. Exim's code has been fixed.
494 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
497 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
499 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
501 PH/37 If a message is not accepted after it has had an id assigned (e.g.
502 because it turns out to be too big or there is a timeout) there is no
503 "Completed" line in the log. When some messages of this type were
504 selected by exigrep, they were listed as "not completed". Others were
505 picked up by some special patterns. I have improved the selection
506 criteria to be more general.
508 PH/38 The host_find_failed option in the manualroute router can now be set
509 to "ignore", to completely ignore a host whose IP address cannot be
510 found. If all hosts are ignored, the behaviour is controlled by the new
511 host_all_ignored option.
513 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
514 homing or because of multiple MX records with /mx) generated more than
515 one IP address, and the following item turned out to be the local host,
516 all the secondary addresses of the first item were incorrectly removed
517 from the list, along with the local host and any following hosts (which
518 is what is supposed to happen).
520 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
521 whoever called Exim into the -H file. In the case of the daemon it was
522 behaving confusingly. When first started, it used values for whoever
523 started the daemon, but after a SIGHUP it used the Exim user (because it
524 calls itself on a restart). I have changed the code so that it now always
527 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
528 message are rejected with the same error (e.g. no authentication or bad
529 sender address), and a DATA command is nevertheless sent (as can happen
530 with PIPELINING or a stupid MUA), the error message that was given to the
531 RCPT commands is included in the rejection of the DATA command. This is
532 intended to be helpful for MUAs that show only the final error to their
535 PH/42 Another patch from the Sieve maintainer.
537 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
538 Eximstats - Fixed some broken HTML links and added missing column headers
540 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
541 columns for Rejects, Temp Rejects, Ham, and Spam rows.
543 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
545 PH/43 Yet another patch from the Sieve maintainer.
547 PH/44 I found a way to check for a TCP/IP connection going away before sending
548 the response to the final '.' that terminates a message, but only in the
549 case where the client has not sent further data following the '.'
550 (unfortunately, this is allowed). However, in many cases there won't be
551 any further data because there won't be any more messages to send. A call
552 to select() can be used: if it shows that the input is "ready", there is
553 either input waiting, or the socket has been closed. An attempt to read
554 the next input character can distinguish the two cases. Previously, Exim
555 would have sent an OK response which the client would never have see.
556 This could lead to message repetition. This fix should cure that, at
557 least in a lot of common cases.
559 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
560 advertised in response to EHLO.
566 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
567 fixed by 4.65/MH/01 (is this a record?) are fixed:
569 (i) An empty string was always treated as zero by the numeric comparison
570 operators. This behaviour has been restored.
572 (ii) It is documented that the numeric comparison operators always treat
573 their arguments as decimal numbers. This was broken in that numbers
574 starting with 0 were being interpreted as octal.
576 While fixing these problems I realized that there was another issue that
577 hadn't been noticed. Values of message_size_limit (both the global option
578 and the transport option) were treated as octal if they started with 0.
579 The documentation was vague. These values are now always treated as
580 decimal, and I will make that clear in the documentation.
586 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
587 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
590 MH/01 Don't check that the operands of numeric comparison operators are
591 integers when their expansion is in "skipping" mode (fixes bug
592 introduced by 4.64-PH/07).
594 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
595 child addresses, Exim now panics and dies. Previously, because the count
596 is held in a short int, deliveries were likely to be lost. As such a
597 large number of recipients for a single message is ridiculous
598 (performance will be very, very poor), I have chosen to impose a limit
599 rather than extend the field.
605 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
606 leftover -K file (the existence of which was triggered by #402).
607 While we were at it, introduced process PID as part of the -K
608 filename. This should rule out race conditions when creating
611 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
612 processing considerably. Previous code took too long for large mails,
613 triggering a timeout which in turn triggers #401.
615 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
616 in the DK code in transports.c. sendfile() is not really portable,
617 hence the _LINUX specificness.
619 TF/01 In the add_headers option to the mail command in an Exim filter,
620 there was a bug that Exim would claim a syntax error in any
621 header after the first one which had an odd number of characters
624 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
625 callout verification, Exim cached a "reject" for the entire domain. This
626 is correct for most verifications, but it is not correct for a recipient
627 verification with use_sender or use_postmaster set, because in that case
628 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
629 case of MAIL FROM:<> rejection from other early rejections (e.g.
630 rejection of HELO). When verifying a recipient using a non-null MAIL
631 address, the cache is ignored if it shows MAIL FROM:<> rejection.
632 Whatever the result of the callout, the value of the domain cache is
633 left unchanged (for any other kind of callout, getting as far as trying
634 RCPT means that the domain itself is ok).
636 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
639 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
640 manifest itself as EPIPE rather than ECONNECT. When tidying away a
641 session, the daemon ignores ECONNECT errors and logs others; it now
642 ignores EPIPE as well.
644 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
645 (quoted-printable decoding).
647 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
648 later the small subsequent patch to fix an introduced bug.
650 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
652 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
654 PH/08 An error is now given if message_size_limit is specified negative.
656 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
657 to be given (somewhat) arbitrary names.
659 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
662 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
663 miscellaneous code fixes
665 PH/10 Added the log_reject_target ACL modifier to specify where to log
668 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
669 hostname. This is wrong, because it relates to the incoming message (and
670 probably the interface on which it is arriving) and not to the outgoing
671 callout (which could be using a different interface). This has been
672 changed to use the value of the helo_data option from the smtp transport
673 instead - this is what is used when a message is actually being sent. If
674 there is no remote transport (possible with a router that sets up host
675 addresses), $smtp_active_hostname is used.
677 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
678 tweaks were necessary in order to get it to work (see also 21 below):
679 (a) The code assumed that strncpy() returns a negative number on buffer
680 overflow, which isn't the case. Replaced with Exim's string_format()
682 (b) There were several signed/unsigned issues. I just did the minimum
683 hacking in of casts. There is scope for a larger refactoring.
684 (c) The code used strcasecmp() which is not a standard C function.
685 Replaced with Exim's strcmpic() function.
686 (d) The code set only $1; it now sets $auth1 as well.
687 (e) A simple test gave the error "authentication client didn't specify
688 service in request". It would seem that Dovecot has changed its
689 interface. Fortunately there's a specification; I followed it and
690 changed what the client sends and it appears to be working now.
692 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
695 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
696 address A is aliased to B and C, where B exists and C does not. Without
697 -v the output is "A verified" because verification stops after a
698 successful redirection if more than one address is generated. However,
699 with -v the child addresses are also verified. Exim was outputting "A
700 failed to verify" and then showing the successful verification for C,
701 with its parentage. It now outputs "B failed to verify", showing B's
702 parentage before showing the successful verification of C.
704 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
705 look up a TXT record in a specific list after matching in a combined
708 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
709 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
710 they consult the DNS. I had assumed they would set it the way they
711 wanted; and indeed my experiments on Linux seem to show that in some
712 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
713 To be on the safe side, however, I have now made the interface to
714 host_find_byname() similar to host_find_bydns(), with an argument
715 containing the DNS resolver options. The host_find_byname() function now
716 sets these options at its start, just as host_find_bydns() does. The smtp
717 transport options dns_qualify_single and dns_search_parents are passed to
718 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
719 of host_find_byname() use the default settings of RES_DEFNAMES
720 (qualify_single) but not RES_DNSRCH (search_parents).
722 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
723 spool_read_header() do less string testing, by means of a preliminary
724 switch on the second character of optional "-foo" lines. (This is
725 overdue, caused by the large number of possibilities that now exist.
726 Originally there were few.) While I was there, I also converted the
727 str(n)cmp tests so they don't re-test the leading "-" and the first
728 character, in the hope this might squeeze out yet more improvement.
730 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
731 flag allowing group syntax was set by the header_syntax check but not
732 turned off, possible causing trouble later; (2) The flag was not being
733 set at all for the header_verify test, causing "group"-style headers to
734 be rejected. I have now set it in this case, and also caused header_
735 verify to ignore an empty address taken from a group. While doing this, I
736 came across some other cases where the code for allowing group syntax
737 while scanning a header line wasn't quite right (mostly, not resetting
738 the flag correctly in the right place). These bugs could have caused
739 trouble for malformed header lines. I hope it is now all correct.
741 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
742 with the "reply" argument non-NULL. The code, however (which originally
743 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
744 but it didn't always do it. This confused somebody who was copying the
745 code for some other use. I have removed all the tests.
747 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
748 feature that was used to support insecure browsers during the U.S. crypto
749 embargo. It requires special client support, and Exim is probably the
750 only MTA that supported it -- and would never use it because real RSA is
751 always available. This code has been removed, because it had the bad
752 effect of slowing Exim down by computing (never used) parameters for the
753 RSA_EXPORT functionality.
755 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
756 authenticator to fail if there's a tab character in the incoming data
757 (there should never be unless someone is messing about, as it's supposed
758 to be base64-encoded). Also added, on Timo's advice, the "secured" option
759 if the connection is using TLS or if the remote IP is the same as the
760 local IP, and the "valid-client-cert option" if a client certificate has
763 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
764 authenticators. This can be used for authorization after authentication
765 succeeds. (In the case of plaintext, it servers for both authentication
768 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
769 if any retry times were supplied.
771 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
772 connection, where there is no client IP address to check. In this
773 situation, the verify now always succeeds.
775 PH/25 Applied John Jetmore's -Mset patch.
777 PH/26 Added -bem to be like -Mset, but loading a message from a file.
779 PH/27 In a string expansion for a processed (not raw) header when multiple
780 headers of the same name were present, leading whitespace was being
781 removed from all of them, but trailing whitespace was being removed only
782 from the last one. Now trailing whitespace is removed from each header
783 before concatenation. Completely empty headers in a concatenation (as
786 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
787 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
789 PH/29 [Removed. This was a change that I later backed out, and forgot to
790 correct the ChangeLog entry (that I had efficiently created) before
791 committing the later change.]
793 PH/30 Exim was sometimes attempting to deliver messages that had suffered
794 address errors (4xx response to RCPT) over the same connection as other
795 messages routed to the same hosts. Such deliveries are always "forced",
796 so retry times are not inspected. This resulted in far too many retries
797 for the affected addresses. The effect occurred only when there were more
798 hosts than the hosts_max_try setting in the smtp transport when it had
799 the 4xx errors. Those hosts that it had tried were not added to the list
800 of hosts for which the message was waiting, so if all were tried, there
801 was no problem. Two fixes have been applied:
803 (i) If there are any address or message errors in an SMTP delivery, none
804 of the hosts (tried or untried) are now added to the list of hosts
805 for which the message is waiting, so the message should not be a
806 candidate for sending over the same connection that was used for a
807 successful delivery of some other message. This seems entirely
808 reasonable: after all the message is NOT "waiting for some host".
809 This is so "obvious" that I'm not sure why it wasn't done
810 previously. Hope I haven't missed anything, but it can't do any
811 harm, as the worst effect is to miss an optimization.
813 (ii) If, despite (i), such a delivery is accidentally attempted, the
814 routing retry time is respected, so at least it doesn't keep
815 hammering the server.
817 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
818 in ${readsocket because some servers need this prod.
820 PH/32 Added some extra debug output when updating a wait-xxx database.
822 PH/33 The hint "could be header name not terminated by colon", which has been
823 given for certain expansion errors for a long time, was not being given
824 for the ${if def:h_colon_omitted{... case.
826 PH/34 The spec says: "With one important exception, whenever a domain list is
827 being scanned, $domain contains the subject domain." There was at least
828 one case where this was not true.
830 PH/35 The error "getsockname() failed: connection reset by peer" was being
831 written to the panic log as well as the main log, but it isn't really
832 panic-worthy as it just means the connection died rather early on. I have
833 removed the panic log writing for the ECONNRESET error when getsockname()
836 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
837 runs only) independently of the message's sender address. This meant
838 that, if the 4xx error was in fact related to the sender, a different
839 message to the same recipient with a different sender could confuse
840 things. In particualar, this can happen when sending to a greylisting
841 server, but other circumstances could also provoke similar problems.
842 I have changed the default so that the retry time for these errors is now
843 based a combination of the sender and recipient addresses. This change
844 can be overridden by setting address_retry_include_sender=false in the
847 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
848 remote server are returned as part of bounce messages. This was not
849 happening for LMTP over a pipe (the lmtp transport), but now it is the
850 same for both kinds of LMTP.
852 PH/38 Despite being documented as not happening, Exim was rewriting addresses
853 in header lines that were in fact CNAMEs. This is no longer the case.
855 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
856 and queue runs started by the daemon processed all messages. This has
857 been fixed so that -R and -S can now usefully be given with -q<time>.
859 PH/40 Import PCRE release 6.7 (fixes some bugs).
861 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
863 PH/42 Give an error if -q is specified more than once.
865 PH/43 Renamed the variables $interface_address and $interface_port as
866 $received_ip_address and $received_port, to make it clear that these
867 values apply to message reception, and not to the outgoing interface when
868 a message is delivered. (The old names remain recognized, of course.)
870 PH/44 There was no timeout on the connect() call when using a Unix domain
871 socket in the ${readsocket expansion. There now is.
873 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
874 be meaningful with "accept".
876 SC/01 Eximstats V1.43
877 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
879 SC/02 Eximstats V1.44
880 Use a glob alias rather than an array ref in the generated
881 parser. This improves both readability and performance.
883 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
884 Collect SpamAssassin and rejection statistics.
885 Don't display local sender or destination tables unless
886 there is data to show.
887 Added average volumes into the top table text output.
889 SC/04 Eximstats V1.46
890 Collect data on the number of addresses (recipients)
891 as well as the number of messages.
893 SC/05 Eximstats V1.47
894 Added 'Message too big' to the list of mail rejection
895 reasons (thanks to Marco Gaiarin).
897 SC/06 Eximstats V1.48
898 Mainlog lines which have GMT offsets and are too short to
899 have a flag are now skipped.
901 SC/07 Eximstats V1.49 (Alain Williams)
902 Added the -emptyok flag.
904 SC/08 Eximstats V1.50
905 Fixes for obtaining the IP address from reject messages.
907 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
908 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
909 whitesspace changes from 4.64-PH/27
911 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
914 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
915 are found, allow negative numbers in numeric criteria)
917 JJ/06 exipick.20061117.2, added new $message_body_missing variable
919 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
920 to match changes made in 4.64-PH/43
922 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
924 PH/47 Put in an explicit test for a DNS lookup of an address record where the
925 "domain" is actually an IP address, and force a failure. This locks out
926 those revolvers/nameservers that support "A-for-A" lookups, in
927 contravention of the specifications.
929 PH/48 When a host name was looked up from an IP address, and the subsequent
930 forward lookup of the name timed out, the host name was left in
931 $sender_host_name, contrary to the specification.
933 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
934 restricted to single-key lookups, Exim was not diagnosing an error if
935 * or *@ was used with a query-style lookup.
937 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
939 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
940 long ago, but noone interested enough thought of it. Let's just say that
941 the "1.1" means that there are some new functions that weren't there at
942 some point in the past.
944 PH/51 Error processing for expansion failure of helo_data from an smtp
945 transport during callout processing was broken.
947 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
948 tested/used via the -bh/-bhc/-bs options.
950 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
951 bug, fixed in subsequent PCRE releases).
953 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
954 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
956 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
962 SC/01 Use a glob alias rather than an array ref in eximstats generated
963 parser. This improves both readability and performance.
965 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
966 Don't display local sender or destination tables in eximstats unless
967 there is data to show.
968 Added average volumes into the eximstats top table text output.
970 SC/03 Collect data on the number of addresses (recipients) as well
971 as the number of messages in eximstats.
973 TF/01 Correct an error in the documentation for the redirect router. Exim
974 does (usually) call initgroups() when daemonizing.
976 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
977 with consistent privilege compared to when running as a daemon.
979 TF/03 Note in the spec that $authenticated_id is not set for local
980 submissions from trusted users.
982 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
983 Thanks to Dean Brooks <dean@iglou.com> for the patch.
985 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
986 by adding some example configuration directives to the default
987 configuration file. A little bit of work is required to uncomment the
988 directives and define how usernames and passwords are checked, but
989 there is now a framework to start from.
991 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
992 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
993 without this. I don't know how relevant this is to other LDAP libraries.
995 PH/02 Add the verb name to the "unknown ACL verb" error.
997 PH/03 Magnus Holmgren's patch for filter_prepend_home.
999 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1001 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1002 directory not expanded when it should be if an expanded home directory
1003 was set for the address (which is overridden by the transport).
1005 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1008 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1009 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1010 because it is too late at that time, and has no effect.
1012 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1013 security issue with \' (bugzilla #107). I could not use the
1014 PQescapeStringConn() function, because it needs a PGconn value as one of
1017 PH/08 When testing addresses using -bt, indicate those final addresses that
1018 are duplicates that would not cause an additional delivery. At least one
1019 person was confused, thinking that -bt output corresponded to deliveries.
1020 (Suppressing duplicates isn't a good idea as you lose the information
1021 about possibly different redirections that led to the duplicates.)
1023 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1024 systems where poll() doesn't work, in particular OS X.
1026 PH/10 Added more information to debugging output for retry time not reached.
1028 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1029 operations in malware.c.
1031 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1034 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1035 syslog_duplication set false, log lines that would normally be written
1036 both the the main log and to the reject log were not written to syslog at
1039 PH/14 In the default configuration, change the use of "message" in ACL warn
1040 statements to "add_header".
1042 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1043 not followed by a command (e.g. "seen endif").
1045 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1046 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1049 PH/17 Added extra conditions to the default value of delay_warning_condition
1053 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1054 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1055 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1058 The Auto-Submitted: and various List- headers are standardised, whereas I
1059 don't think Precedence: ever was.
1061 PH/18 Refactored debugging code in route_finduser() to show more information,
1062 in particular, the error code if getpwnam() issues one.
1064 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1065 This is apparently needed in addition to the PH/07 change above to avoid
1066 any possible encoding problems.
1068 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1069 but not after initializing Perl.
1071 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1072 output them only if debugging. By default they are written stderr,
1073 apparently, which is not desirable.
1075 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1078 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1081 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1083 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1084 authenticated or an ident call has been made. Suppress the default
1085 values for $authenticated_id and $authenticated_sender (but permit -oMai
1086 and -oMas) when testing with -bh.
1088 PH/24 Re-jigged the order of the tests in the default configuration so that the
1089 tests for valid domains and recipients precede the DNS black list and CSA
1090 tests, on the grounds that those ones are more expensive.
1092 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1093 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1094 This bug exists in every version of Exim that I still have, right back to
1097 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1098 However, an attempt to turn on case-sensitivity in a regex key by
1099 including (?-i) didn't work because the subject string was already
1100 lowercased, and the effects were non-intuitive. It turns out that a
1101 one-line patch can be used to allow (?-i) to work as expected.
1107 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1108 other effects) broke the use of negated acl sub-conditions.
1110 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1113 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1114 "Deny" causes Exim to reject the incoming connection with a 554 error.
1115 Unfortunately, if there is a major crisis, such as a disk failure,
1116 tcp-wrappers gives "deny", whereas what one would like would be some
1117 kind of temporary error. A kludge has been added to help with this.
1118 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1119 554 error is used if errno is still zero or contains ENOENT (which occurs
1120 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1123 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1125 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1126 errors. Otherwise a message that provokes a temporary error (when other
1127 messages do not) can cause a whole host to time out.
1129 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1130 addresses were routed directly to files or pipes from a redirect router.
1131 File deliveries just didn't batch; pipe deliveries might have suffered
1134 PH/06 A failure to get a lock for a hints database would erroneously always say
1135 "Failed to get write lock", even when it was really a read lock.
1137 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1138 of 0600. This has been changed to use the value of the lockfile_mode
1139 option (which defaults to 0600).
1141 PH/08 Applied small patch from the Sieve maintainer.
1143 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1144 folder from quota calculations, a direct delivery into this folder messed
1145 up the contents of the maildirsize file. This was because the regex was
1146 used only to exclude .Trash (or whatever) when the size of the mailbox
1147 was calculated. There was no check that a delivery was happening into an
1148 excluded directory. This bug has been fixed by ignoring all quota
1149 processing for deliveries into excluded directories.
1151 PH/10 Added the maildirfolder_create_regex option to appendfile.
1157 PH/01 The code for finding all the local interface addresses on a FreeBSD
1158 system running IPv6 was broken. This may well have applied to all BSD
1159 systems, as well as to others that have similar system calls. The broken
1160 code found IPv4 interfaces correctly, but gave incorrect values for the
1161 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1162 that it would not match correctly against @[] and not recognize the IPv6
1165 PH/02 The ipliteral router was not recognizing addresses of the form user@
1166 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1168 PH/03 Added disable_ipv6.
1170 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1171 decoded form, because it is most often used to construct To: headers
1172 lines in autoreplies, and the decoded form may well be syntactically
1173 invalid. However, $reply_address has leading white space removed, and all
1174 newlines turned into spaces so that the autoreply transport does not
1177 PH/05 If group was specified without a user on a router, and no group or user
1178 was specified on a transport, the group from the router was ignored.
1180 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1181 for visible compile-time settings that can be used to change these
1182 numbers, for those that want even more. Backwards compatibility with old
1183 spool files has been maintained. However, going back to a previous Exim
1184 release will lost any variables that are in spool files.
1186 PH/07 Two small changes when running in the test harness: increase delay when
1187 passing a TCP/IP connection to a new process, in case the original
1188 process has to generate a bounce, and remove special handling of
1189 127.0.0.2 (sic), which is no longer necessary.
1191 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1192 be the same on different OS.
1194 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1197 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1198 whether --show-vars was specified or not
1200 JJ/02 exipick: Added support for new ACL variable spool format introduced
1203 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1204 syntactically invalid From: or Reply-to: line, and a filter used this to
1205 generate an autoreply, and therefore failed to obtain an address for the
1206 autoreply, Exim could try to deliver to a non-existent relative file
1207 name, causing unrelated and misleading errors. What now happens is that
1208 it logs this as a hard delivery error, but does not attempt to create a
1211 PH/11 The exinext utility has a -C option for testing purposes, but although
1212 the given file was scanned by exinext itself; it wasn't being passed on
1213 when Exim was called.
1215 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1216 an end-of-file indication when reading a command response.
1218 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1219 compiled. In many other places in Exim, IPv6 addresses are always
1220 recognized, so I have changed this. It also means that IPv4 domain
1221 literals of the form [IPV4:n.n.n.n] are now always recognized.
1223 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1224 used if the router is not running as root, for example, when verifying at
1225 ACL time, or when using -bh. The debugging output from this situation was
1226 non-existent - all you got was a failure to exec. I have made two
1229 (a) Failures to set uid/gid, the current directory, or a process leader
1230 in a subprocess such as that created by queryprogram now generate
1231 suitable debugging ouput when -d is set.
1233 (b) The queryprogram router detects when it is not running as root,
1234 outputs suitable debugging information if -d is set, and then runs
1235 the subprocess without attempting to change uid/gid.
1237 PH/15 Minor change to Makefile for building test_host (undocumented testing
1240 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1241 additional section of a DNS packet that returns MX or SRV records.
1242 Instead, it always explicitly searches for A/AAAA records. This avoids
1243 major problems that occur when a DNS server includes only records of one
1244 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1245 fixed another bug: if SRV records were looked up and the corresponding
1246 address records were *not* found in the additional section, the port
1247 values from the SRV records were lost.
1249 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1250 using the correct key (the original address) when searching the retry
1251 rules in order to find which one to use for generating the retry hint.
1253 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1254 adding the default one. Similarly, if it contains a Reply-To: header, the
1255 errors_reply_to option, if set, is not used.
1257 PH/19 When calculating a retry time, Exim used to measure the "time since
1258 failure" by looking at the "first failed" field in the retry record. Now
1259 it does not use this if it is later than than the arrival time of the
1260 message. Instead it uses the arrival time. This makes for better
1261 behaviour in cases where some deliveries succeed, thus re-setting the
1262 "first failed" field. An example is a quota failure for a huge message
1263 when small messages continue to be delivered. Without this change, the
1264 "time since failure" will always be short, possible causing more frequent
1265 delivery attempts for the huge message than are intended.
1266 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1268 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1269 $1, $2, $3) because the numerical variables can be reset during some
1270 expansion items (e.g. "match"), thereby losing the authentication data.
1272 PH/21 Make -bV show the size of off_t variables so that the test suite can
1273 decide whether to run tests for quotas > 2G.
1275 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1276 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1277 filecount value is greater than 2G or if a quota value is greater than 2G
1278 on a system where the size of off_t is not greater than 4, a panic error
1281 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1282 never match. The debug and -bh output now contains an explicit error
1283 message indicating a malformed IPv4 address or mask.
1285 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1286 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1287 PH/23 above applies.
1289 PH/25 Do not write to syslog when running in the test harness. The only
1290 occasion when this arises is a failure to open the main or panic logs
1291 (for which there is an explicit test).
1293 PH/26 Added the /no_tell option to "control=freeze".
1295 PH/27 If a host name lookup failed very early in a connection, for example, if
1296 the IP address matched host_lookup and the reverse lookup yielded a name
1297 that did not have a forward lookup, an error message of the form "no IP
1298 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1299 could be logged. Now it outputs the IP address instead of "NULL".
1301 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1302 allows the sender and the authenticated sender to be set when
1303 submitting a message from within Exim. Since child_open_exim() is
1304 documented for local_scan(), the new function should be too.
1306 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1307 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1308 results in an empty string is now treated as unset.
1310 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1312 PH/31 Added sender_verify_fail logging option.
1314 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1315 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1316 tidied the source and removed it altogether.
1318 PH/33 When a queue run was abandoned because the load average was too high, a
1319 log line was always written; now it is written only if the queue_run log
1320 selector is set. In addition, the log line for abandonment now contains
1321 information about the queue run such as the pid. This is always present
1322 in "start" and "stop" lines but was omitted from the "abandon" line.
1324 PH/34 Omit spaces between a header name and the colon in the error message that
1325 is given when verify = headers_syntax fails (if there are lots of them,
1326 the message gets confusing).
1328 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1329 names, as there are now some PTR records that contain slashes. This check
1330 is only to protect against broken name servers that fall over on strange
1331 characters, so the fact that it applies to all lookups doesn't matter.
1333 PH/36 Now that the new test suite is complete, we can remove some of the
1334 special code in Exim that was needed for the old test suite. For example,
1335 sorting DNS records because real resolvers return them in an arbitrary
1336 order. The new test suite's fake resolver always returns records in the
1339 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1340 bounces) except when queue_only is set, to avoid logging races between
1341 the different processes.
1343 PH/38 Panic-die if .include specifies a non-absolute path.
1345 PH/39 A tweak to the "H" retry rule from its user.
1347 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1348 a label. They prevented compilation on older perls.
1350 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1351 a warning to be raised on newish perls.
1353 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1354 on queue. Changes to match documented behaviour of showing count of
1355 messages matching specified criteria.
1357 PH/40 Changed the default ident timeout from 30s to 5s.
1359 PH/41 Added support for the use of login_cap features, on those BSD systems
1360 that have them, for controlling the resources used by pipe deliveries.
1362 PH/42 The content-scanning code uses fopen() to create files in which to put
1363 message data. Previously it was not paying any attention to the mode of
1364 the files. Exim runs with umask(0) because the rest of the code creates
1365 files with open(), and sets the required mode explicitly. Thus, these
1366 files were ending up world-writeable. This was not a big issue, because,
1367 being within the spool directory, they were not world-accessible. I have
1368 created a function called modefopen, which takes an additional mode
1369 argument. It sets umask(777), creates the file, chmods it to the required
1370 mode, then resets the umask. All the relevant calls to fopen() in the
1371 content scanning code have been changed to use this function.
1373 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1374 to 24 hours. This avoids potential overflow problems when processing G
1375 and H retry rules. I suspect nobody ever tinkers with this value.
1377 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1379 PH/45 When the plaintext authenticator is running as a client, the server's
1380 challenges are checked to ensure they are valid base64 strings. By
1381 default, the authentication attempt is cancelled if an invalid string is
1382 received. Setting client_ignore_invalid_base64 true ignores these errors.
1383 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1384 they are received. Thus, the responses can be made to depend on the
1385 challenges. If an invalid string is ignored, an empty string is placed in
1388 PH/46 Messages that are created by the autoreply transport now contains a
1389 References: header, in accordance with RFCs 2822 and 3834.
1391 PH/47 Added authenticated_sender_force to the smtp transport.
1393 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1395 PH/49 Installed latest patch from the Sieve maintainer.
1397 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1398 was also set, the appendfile transport was unnecessarily scanning a
1399 directory of message files (e.g. for maildir delivery) to find the count
1400 of files (along with the size), even though it did not need this
1401 information. It now does the scan only if it needs to find either the
1402 size of the count of files.
1404 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1406 PH/52 Two bugs concerned with error handling when the smtp transport is
1409 (i) Exim was not creating retry information for temporary errors given
1410 for individual recipients after the DATA command when the smtp transport
1411 was used in LMTP mode. This meant that they could be retried too
1412 frequently, and not timed out correctly.
1414 (ii) Exim was setting the flag that allows error details to be returned
1415 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1416 recipients that were returned after the DATA command.
1418 PH/53 This is related to PH/52, but is more general: for any failing address,
1419 when detailed error information was permitted to be returned to the
1420 sender, but the error was temporary, then after the final timeout, only
1421 "retry timeout exceeded" was returned. Now it returns the full error as
1422 well as "retry timeout exceeded".
1424 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1425 do this, and (what is worse) MTAs that accept it.
1427 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1428 will now be deprecated.
1430 PH/56 New os.c-cygwin from the Cygwin maintainer.
1432 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1433 formats (previously only available in exim formats via -bpr, -bpru,
1434 and -bpra. Now also available in native and exiqgrep formats)
1436 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1437 with very large, slow to parse queues
1439 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1441 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1443 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1444 responses to authentication challenges, though it was showing the
1445 challenges; (ii) I've removed the CR characters from the debug output for
1448 PH/58 Allow for the insertion of a newline as well as a space when a string
1449 is turned into more than one encoded-word during RFC 2047 encoding. The
1450 Sieve code now uses this.
1452 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1453 data_4xx, lost_connection, tls_required.
1455 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1456 message was being sent as an SMTP response.
1458 PH/61 Add -l and -k options to exicyclog.
1460 PH/62 When verifying, if an address was redirected to one new address, so that
1461 verification continued, and the new address failed or deferred after
1462 having set something in $address_data, the value of $address_data was not
1463 passed back to the ACL. This was different to the case when no
1464 redirection occurred. The value is now passed back in both cases.
1466 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1467 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1468 use login_cap.h, so on its own it isn't the distinguishing feature. The
1469 new name refers directly to the setclassresources() function.
1471 PH/65 Added configuration files for NetBSD3.
1473 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1475 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1476 is preferred over IPv4.
1478 PH/68 The bounce_return_message and bounce_return_body options were not being
1479 honoured for bounces generated during the reception of non-SMTP messages.
1480 In particular, this applied to messages rejected by the ACL. This bug has
1481 been fixed. However, if bounce_return_message is true and bounce_return_
1482 body is false, the headers that are returned for a non-SMTP message
1483 include only those that have been read before the error was detected.
1484 (In the case of an ACL rejection, they have all been read.)
1486 PH/69 The HTML version of the specification is now built in a directory called
1487 spec_html instead of spec.html, because the latter looks like a path with
1488 a MIME-type, and this confuses some software.
1490 PH/70 Catch two compiler warnings in sieve.c.
1492 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1493 function verify_get_ident() calls ip_connect() to connect a socket, but
1494 if the "connect()" function timed out, ip_connect() used to close the
1495 socket. However, verify_get_ident() also closes the socket later, and in
1496 between Exim writes to the log, which may get opened at this point. When
1497 the socket was closed in ip_connect(), the log could get the same file
1498 descriptor number as the socket. This naturally causes chaos. The fix is
1499 not to close the socket in ip_connect(); the socket should be closed by
1500 the function that creates it. There was only one place in the code where
1501 this was missing, in the iplookup router, which I don't think anybody now
1502 uses, but I've fixed it anyway.
1504 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1505 well as to direct DNS lookups. Otherwise the handling of names in host
1506 lists is inconsistent and therefore confusing.
1512 PH/01 Two changes to the default runtime configuration:
1514 (1) Move the checks for relay_from_hosts and authenticated clients from
1515 after to before the (commented out) DNS black list checks.
1517 (2) Add control=submission to the relay_from_hosts and authenticated
1518 clients checks, on the grounds that messages accepted by these
1519 statements are most likely to be submissions.
1521 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1523 (1) Generate an error if the third argument for the ${prvs expansion is
1526 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1529 (3) Reset the variables that are obtained from the first argument of
1530 ${prvscheck and used in the second argument before leaving the code,
1531 because their memory is reclaimed, so using them afterwards may do
1534 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1535 one (it's much easier than Tom thought :-).
1537 (5) Because of (4), we can now allow for the use of $prvscheck_result
1538 inside the third argument.
1540 PH/03 For some reason, the default setting of PATH when running a command from
1541 a pipe transport was just "/usr/bin". I have changed it to
1544 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1545 anything to be listed in the output from -bV.
1547 PH/05 When a filter generated an autoreply, the entire To: header line was
1548 quoted in the delivery log line, like this:
1550 => >A.N.Other <ano@some.domain> <original@ddress> ...
1552 This has been changed so that it extracts the operative address. There
1553 may be more than one such address. If so, they are comma-separated, like
1556 => >ano@some.domain,ona@other.domain <original@ddress> ...
1558 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1559 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1560 not being looked up in the rDNS to get a host name, Exim was showing the
1561 IP address twice in Received: lines, even though the IP addresses were
1562 identical. For example:
1564 Received: from [1.2.3.4] (helo=[1.2.3.4])
1566 However, if the real host name was known, it was omitting the HELO data
1567 if it matched the actual IP address. This has been tidied up so that it
1568 doesn't show the same IP address twice.
1570 PH/07 When both +timestamp and +memory debugging was on, the value given by
1571 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1572 called by the debug printing, thereby overwriting the timestamp buffer.
1573 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1575 PH/08 When the original message was included in an autoreply transport, it
1576 always said "this is a copy of the message, including all the headers",
1577 even if body_only or headers_only was set. It now gives an appropriate
1580 PH/09 Applied a patch from the Sieve maintainer which:
1582 o fixes some comments
1583 o adds the (disabled) notify extension core
1584 o adds some debug output for the result of if/elsif tests
1585 o points to the current vacation draft in the documentation
1586 and documents the missing references header update
1590 o fixes a bug in processing the envelope test (when testing
1591 multiple envelope elements, the last element determinted the
1594 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1595 Electronic Mail") by including:
1597 Auto-submitted: auto-generated
1599 in the messages that it generates (bounce messages and others, such as
1600 warnings). In the case of bounce messages for non-SMTP mesages, there was
1601 also a typo: it was using "Auto_submitted" (underscore instead of
1602 hyphen). Since every message generated by Exim is necessarily in response
1603 to another message, thes have all been changed to:
1605 Auto-Submitted: auto-replied
1607 in accordance with these statements in the RFC:
1609 The auto-replied keyword:
1611 - SHOULD be used on messages sent in direct response to another
1612 message by an automatic process,
1614 - MUST NOT be used on manually-generated messages,
1616 - MAY be used on Delivery Status Notifications (DSNs) and Message
1617 Disposition Notifications (MDNs),
1619 - MUST NOT be used on messages generated by automatic or periodic
1620 processes, except for messages which are automatic responses to
1623 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1624 to the default Received: header definition.
1626 PH/12 Added log selector acl_warn_skipped (default on).
1628 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1629 variables because (a) they are in the wrong storage pool and (b) even if
1630 they were copied, it wouldn't work properly because of the caching.
1632 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1633 checking when decoding. Apparently there are clients that generate
1634 overlong encoded strings. Why am I not surprised?
1636 PH/15 If the first argument of "${if match_address" was not empty, but did not
1637 contain an "@" character, Exim crashed. Now it writes a panic log message
1638 and treats the condition as false.
1640 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1642 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1643 extension "envelope-auth". The code is finished and in agreement with
1644 other implementations, but there is no documentation so far and in fact,
1645 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1646 not changing the active code.
1648 Print executed "if" and "elsif" statements when debugging is used. This
1649 helps a great deal to understand what a filter does.
1651 Document more things not specified clearly in RFC3028. I had all this
1652 sorted out, when out of a sudden new issues came to my mind. Oops."
1654 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1657 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1658 canonical form (as documented). However, after a host name lookup from
1659 the IP address, check_host() was doing a simple string comparison with
1660 addresses acquired from the DNS when checking that the found name did
1661 have the original IP as one of its addresses. Since any found IPv6
1662 addresses are likely to be in abbreviated form, the comparison could
1663 fail. Luckily, there already exists a function for doing the comparison
1664 by converting both addresses to binary, so now that is used instead of
1665 the text comparison.
1667 PH/20 There was another similar case to PH/19, when a complete host name was
1668 given in a host list; looking up its IP address could give an abbreviated
1669 form, whereas the current host's name might or might not be abbreviated.
1670 The same fix has been applied.
1676 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1677 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1680 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1681 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1683 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1685 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1686 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1687 Also, for "Auto-Submitted", treat anything other than "no" as
1688 non-personal, in accordance with RFC 3834. (Previously it treated
1689 anything starting "auto-" as non-personal.)
1691 TF/01 The control=submission/name=... option had a problem with syntax
1692 errors if the name included a slash character. The /name= option
1693 now slurps the rest of the string, so it can include any characters
1694 but it must come last in the list of options (after /sender_retain
1697 PH/05 Some modifications to the interface to the fake nameserver for the new
1705 TK/01 Added the "success_on_redirect" address verification option. See
1706 NewStuff for rationale and an example.
1708 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1710 PH/02 Patch to exigrep to allow it to work on syslog lines.
1712 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1713 fread() to skip over the body file's header line, because in Cygwin the
1714 header line is locked and is inaccessible.
1716 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1717 co-exist for some time) to make it clear that it is the Exim ID that is
1718 referenced, not the Message-ID: header line.
1720 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1721 string_format() function, because snprintf() does not exist on all
1724 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1725 ${stat: expansion item.
1727 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1728 protocol synchronization error", to keep the pedants happy.
1730 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1731 well as for IRIX systems, when gcc is being used. See the host.c source
1734 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1736 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1739 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1740 transport and to the smtp transport in LMTP mode.
1742 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1744 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1745 run a filter in a subprocess. This could lead to confusion in subsequent
1746 lookups in the parent process. There should also be a search_tidyup() at
1747 the end of the subprocess.
1749 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1750 only if the host matched helo_try_verify_hosts, which caused the
1751 verification to occur when the EHLO/HELO command was issued. The ACL just
1752 tested the remembered result. Now, if a previous verification attempt has
1753 not happened, "verify = helo" does it there and then.
1755 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1757 TK/03 Fix log output including CR from clamd.
1759 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1760 exist provoked a memory error which could cause a segfault.
1762 PH/15 Installed PCRE 6.2
1764 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1766 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1767 of the problem. Specifically, suggested +O2 rather than +O1 for the
1770 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1772 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1773 record, the port was not correctly logged when the outgoing_port log
1774 selector was set (it logged the transort's default port).
1776 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1777 fallback_hosts, and "hosts" in the smtp transport.
1779 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1780 host errors such as "Connection refused".
1782 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1783 authentication with radiusclient 0.4.9:
1785 - Error returned from rc_read_config was caught wrongly
1786 - Username/password not passed on to radius server due to wrong length.
1788 The presumption is that some radiusclient API changes for 4.51/PH/17
1789 were not taken care of correctly. The code is still untested by me (my
1790 Linux distribution still has 0.3.2 of radiusclient), but it was
1791 contributed by a Radius user.
1793 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1794 expanding the "port" option of the smtp transport.
1796 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1797 while reading a MIME header. Thanks to Tom Hughes for a patch.
1799 PH/24 Include config.h inside local_scan.h so that configuration settings are
1802 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1803 that in an ACL for RCPT (for example), you can examine exactly what was
1806 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1807 commands, but it was not correctly comparing the address with the actual
1808 client host address. Thus, it would show the EHLO address in Received:
1809 header lines when this was not necessary.
1811 PH/27 Added the % operator to ${eval:}.
1813 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1814 it should be ignoring failures (because with -C, for example, it has lost
1815 privilege). It wasn't ignoring creation failures other than "already
1818 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1821 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1822 because an input line was too long, either on its own, or by virtue of
1823 too many continuations, the temporary file was not being removed, and the
1824 return code was incorrect.
1826 PH/31 Missing "BOOL" in function definition in filtertest.c.
1828 PH/32 Applied Sieve patches from the maintainer.
1830 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1832 PH/33 Added "verify = not_blind".
1834 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1835 Local/Makefile (with some defaults set). These are used in built scripts
1836 such as exicyclog, but they have never been used in the exim_install
1837 script (though there are many overriding facilities there). I have
1838 arranged that the exim_install script now takes note of these two
1841 PH/35 Installed configuration files for Dragonfly.
1843 PH/36 When a locally submitted message by a trusted user did not contain a
1844 From: header, and the sender address was obtained from -f or from an SMTP
1845 MAIL command, and the trusted user did not use -F to supply a sender
1846 name, $originator_name was incorrectly used when constructing a From:
1847 header. Furthermore, $originator_name was used for submission mode
1848 messages from external hosts without From: headers in a similar way,
1849 which is clearly wrong.
1851 PH/37 Added control=suppress_local_fixups.
1853 PH/38 When log_selector = +received_sender was set, and the addition of the
1854 sender made the log line's construction buffer exactly full, or one byte
1855 less than full, an overflow happened when the terminating "\n" was
1858 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1859 when the result of a list match is failure because a DNS lookup failed.
1861 PH/40 RM_COMMAND is now used in the building process.
1863 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1864 the "build-* directories that it finds.
1866 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1867 address in a domain literal was a prefix of an interface address.
1869 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1870 when verifying a sender address, unless rewrite_headers is false.
1872 PH/44 Wrote a long comment about why errors_to addresses are verified as
1873 recipients, not senders.
1875 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1876 the ratelimit ACL was added.
1878 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1880 PH/46 Added extra information about PostgreSQL errors to the error string.
1882 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1883 avoiding the need to install special zones in a real server. This is
1884 backwards compatible; if it can't find the fake resolver, it drops back.
1885 Thus, both old and new test suites can be run.
1887 TF/02 Added util/ratelimit.pl
1889 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1890 clock is set back in time.
1892 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1893 Candler <B.Candler@pobox.com>.
1895 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1896 OK for addresses that are the result of redirections.
1898 PH/48 A number of further additions for the benefit of the new test suite,
1899 including a fake gethostbyname() that interfaces to the fake DNS resolver
1902 TF/06 The fix for widen_domains has also been applied to qualify_single and
1903 search_parents which are the other dnslookup options that can cause
1906 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1909 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1911 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1912 service" if it encountered an MX record with an empty target hostname.
1913 The message is now "an MX or SRV record indicated no SMTP service".
1915 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1916 if the DNS of the sending site is misconfigured. This is quite a
1917 common situation. This change restores the behaviour of treating a
1918 helo verification defer as a failure.
1920 PH/51 If self=fail was set on a router, the bounce message did not include the
1921 actual error message.
1927 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1929 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1930 command itself ended in error, the underlying message about the transport
1931 filter timeout was being overwritten with the pipe command error. Now the
1932 underlying error message should be appended to the second error message.
1934 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1936 PH/02 Reduce the amount of output that "make" produces by default. Full output
1937 can still be requested.
1939 PH/03 The warning log line about a condition test deferring for a "warn" verb
1940 was being output only once per connection, rather than after each
1941 occurrence (because it was using the same function as for successful
1942 "warn" verbs). This seems wrong, so I have changed it.
1944 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1945 it should not have, which might have caused a crash in the right
1946 circumstances, but probably never did.
1948 PH/04 Installed a modified version of Tony Finch's patch to make submission
1949 mode fix the return path as well as the Sender: header line, and to
1950 add a /name= option so that you can make the user's friendly name appear
1953 TF/03 Added the control = fakedefer ACL modifier.
1955 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1956 Mark Lowes for thorough testing.
1958 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1960 TK/03 Merged latest SRS patch from Miles Wilton.
1962 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1963 with the definition in sysexits.h (which is #included earlier).
1964 Fortunately, Exim does not actually use EX_OK. The code used to try to
1965 preserve the sysexits.h value, by assumimg that macro definitions were
1966 scanned for macro replacements. I have been disabused of this notion,
1967 so now the code just undefines EX_OK before #including unistd.h.
1969 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1970 in the smtp transport. When a block could not be written in a single
1971 write() function, the timeout was being re-applied to each part-write.
1972 This seems wrong - if the receiver was accepting one byte at a time it
1973 would take for ever. The timeout is now adjusted when this happens. It
1974 doesn't have to be particularly precise.
1976 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1977 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1979 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1980 without a domain if the check to <postmaster@domain> fails.
1982 SC/01 Eximstats: added -xls and the ability to specify output files
1983 (patch written by Frank Heydlauf).
1985 SC/02 Eximstats: use FileHandles for outputing results.
1987 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1989 SC/04 Eximstats: fixed display of large numbers with -nvr option
1991 SC/05 Eximstats: fixed merging of reports with empty tables.
1993 SC/06 Eximstats: added the -include_original_destination flag
1995 SC/07 Eximstats: removed tabs and trailing whitespace.
1997 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1999 TK/06 MBOX spool code: Add real "From " MBOX separator line
2000 so the .eml file is really in mbox format (even though
2001 most programs do not really care). Patch from Alex Miller.
2003 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2004 The latter is generated from $received_to and is only set if the
2005 message has one envelope recipient. SA can use these headers,
2006 obviously out-of-the-box. Patch from Alex Miller.
2008 PH/08 The ${def test on a variable was returning false if the variable's
2009 value was "0", contrary to what the specification has always said!
2010 The result should be true unless the variable is empty.
2012 PH/09 The syntax error of a character other than { following "${if
2013 def:variable_name" (after optional whitespace) was not being diagnosed.
2014 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2015 accidental colon was present, for example, could give incorrect results.
2017 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2018 result is used (not including appendfile, where other changes are about
2021 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2022 This involved changing a lot of size variables from int to off_t. It
2023 should work with maildirs and everything.
2025 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2026 spamd dying while we are connected to it.
2028 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2031 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2032 function for BDB 4.3.
2034 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2036 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2037 the sources for PCRE have been re-arranged and more files are now
2040 PH/15 The code I had for printing potentially long long variables in PH/11
2041 above was not the best (it lost precision). The length of off_t variables
2042 is now inspected at build time, and an appropriate printing format (%ld
2043 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2044 to be "long long int" or "long int". This is needed for the internal
2045 formatting function string_vformat().
2047 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2048 the configuration file to be ":syslog", then the script "guesses" where
2049 the logs files are, rather than using the compiled in default. In our
2050 case the guess is not the same as the compiled default, so the script
2051 suddenly stopped working when I started to use syslog. The patch checks
2052 to see if log_file_path is "". If so, it attempts to read it from exim
2053 with no configuration file to get the compiled in version, before it
2054 falls back to the previous guessing code."
2056 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2057 implementing BATV in an Exim configuration. See NewStuff for the gory
2060 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2061 Makefile that are specific to HP-UX.
2063 PH/18 If the "use_postmaster" option was set for a recipient callout together
2064 with the "random" option, the postmaster address was used as the MAIL
2065 FROM address for the random test, but not for the subsequent recipient
2066 test. It is now used for both.
2068 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2069 patch removes a few documentation additions to RFC 3028, because the
2070 latest draft now contains them. It adds the new en;ascii-case comparator
2071 and a new error check for 8bit text in MIME parts. Comparator and
2072 require names are now matched exactly. I enabled the subaddress
2073 extension, but it is not well tested yet (read: it works for me)."
2075 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2076 rework some of the code of TK/09 above to avoid the hardwired use of
2077 "%lld" and "long long". Replaced the call to snprintf() with a call to
2080 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2081 records point to non-existent hosts", "retry timeout exceeded", and
2082 "retry time not reached for any host after a long failure period".
2084 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2085 experimental DomainKeys support:
2087 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2088 (2) On an error such as an illegally used "control", the wrong name for
2089 the control was given.
2091 These problems did NOT occur unless DomainKeys support was compiled.
2093 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2095 PH/24 Added ${if match_ip condition.
2097 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2098 databases so that it will be absolutely obvious if a crash occurs in the
2099 DB library. This is a regular occurrence (often caused by mis-matched
2102 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2103 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2104 picked up on a user's system that detects such things. There doesn't seem
2105 to be a gcc warning option for this - only an attribute that has to be
2106 put on the function's prototype. It seems that in Fedora Core 4 they have
2107 set this on a number of new functions. No doubt there will be more in due
2110 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2111 specify a transport. However, if an address that was verified by such a
2112 router was the subject of a callout, Exim crashed because it tried to
2113 read the rcpt_include_affixes from the non-existent transport. Now it
2114 just assumes that the setting of that option is false. This bug was
2115 introduced by 4.51/PH/31.
2117 PH/28 Changed -d+all to exclude +memory, because that information is very
2118 rarely of interest, but it makes the output a lot bigger. People tend to
2119 do -d+all out of habit.
2121 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2122 code in os-type was giving problems when libc.so lives in lib64, like on
2125 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2126 aren't the modern standard, and it seems that some systems' include files
2127 don't always have them. Exim was already checking for some of the newer
2128 ones like T_AAAA, and defining it itself. I've added checks for all the
2129 record types that Exim uses.
2131 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2132 not automatically generating a new one, as it is supposed to. This
2133 prevented TLS from working. If the file did exist, but contained invalid
2134 data, a new version was generated, as expected. It was only the case of a
2135 non-existent file that was broken.
2137 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2138 with a change in libdomainkeys > 0.64.
2140 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2141 from DNS. If the selector record carries the flag, it now has
2142 precedence over the domain-wide flag.
2144 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2146 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2147 the use of a port name that isn't defined in /etc/services) occurred, the
2148 message was deferred as in a normal delivery, and thus remained on the
2149 spool, instead of being failed because of the mua_wrapper setting. This
2150 is now fixed, and I tidied up some of the mua_wrapper messages at the
2153 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2154 the messages in a hash of arrays rather than using individual hashes.
2155 This is a bit cleaner and results in dramatic memory savings, albeit
2156 at a slight CPU cost.
2158 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2159 as requested by Marc Sherman.
2161 SC/10 Eximstats: added histograms for user specified patterns as requested
2164 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2166 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2167 fopen() in the content-scanning modules that did not already have it.
2173 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2174 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2176 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2178 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2180 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2181 to test Sieve filters that use "vacation".
2183 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2184 that changes the way the GnuTLS parameters are stored in the cache file.
2185 The new format can be generated externally. For backward compatibility,
2186 if the data in the cache doesn't make sense, Exim assumes it has read an
2187 old-format file, and it generates new data and writes a new file. This
2188 means that you can't go back to an older release without removing the
2191 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2192 work if there are any delivery delays because "one_time" forces the
2193 parent to be marked "delivered", so its unseen clone is never tried
2194 again. For this reason, Exim now forbids the simultaneous setting of
2197 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2198 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2199 entry if you want to know the details. The fix, however, neglected to
2200 consider the case when local delivery batching is involved. The test for
2201 "previously delivered" was not happening when checking to see if an
2202 address could be batched with a previous (undelivered) one; under
2203 certain circumstances this could lead to multiple deliveries to the same
2206 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2207 in its include files, and this causes problems building Exim.
2209 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2210 header_syntax) but Exim was just ignoring anything given after a slash.
2211 In particular, this caused confusion with an attempt to use "verify =
2212 reverse_host_lookup/defer_ok". An error is now given when options are
2213 supplied for verify items that do not have them. (Maybe reverse_host_
2214 lookup should have a defer_ok option, but that's a different point.)
2216 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2217 defined by RFC 821) to 2048, because there were problems with some AUTH
2218 commands, and RFC 1869 says the size should be increased for extended
2219 SMTP commands that take arguments.
2221 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2224 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2225 "unknown" error; now it says that the functionality isn't in the binary.
2227 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2228 an address' error message when a string expansion fails (syntax or
2229 whatever). Otherwise the password may appear in the log. Following change
2230 PH/42 below, there is no longer a chance of it appearing in a bounce
2233 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2235 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2236 of its addresses was ever tried. (Bugzilla bug #2.)
2238 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2239 the result incorrectly in the debug output. (It correctly added a newline
2240 to what was transported.)
2242 TF/01 Added $received_time.
2244 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2245 commented out examples of how to interface to a virus scanner and to
2246 SpamAssassin. Also added commented examples of av_scanner and
2247 spamd_address settings.
2249 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2250 and controls are allowed in which ACLs. There were a couple of minor
2251 errors. Some of the entries in the conditions table (which is a table of
2252 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2253 negation of where the condition IS allowed.
2255 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2257 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2258 header file does not have a version number, so I've had to invent a new
2259 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2260 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2261 radiusclient), but it was contributed by a Radius user.
2263 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2264 files or directories, for OpenSSL.
2266 PH/19 When an Exim process that is running as root has to create an Exim log
2267 file, it does so in a subprocess that runs as exim:exim so as to get the
2268 ownership right at creation (otherwise, other Exim processes might see
2269 the file with the wrong ownership). There was no test for failure of this
2270 fork() call, which would lead to the process getting stuck as it waited
2271 for a non-existent subprocess. Forks do occasionally fail when resources
2272 run out. I reviewed all the other calls to fork(); they all seem to check
2275 PH/20 When checking for unexpected SMTP input at connect time (before writing
2276 the banner), Exim was not dealing correctly with a non-positive return
2277 from the read() function. If the client had disconnected by this time,
2278 the result was a log entry for a synchronization error with an empty
2279 string after "input=" when read() returned zero. If read() returned -1
2280 (an event I could not check), uninitialized data bytes were printed.
2281 There were reports of junk text (parts of files, etc) appearing after
2284 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2286 PH/22 Added support for macro redefinition, and (re)definition in between
2287 driver and ACL definitions.
2289 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2290 forgetting to use the resulting value; it was using the unexpanded value.
2292 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2293 hadn't been configured. The fix is from Juergen Kreileder, who
2294 understands it better than I do:
2296 "Here's what I see happening with three configured cyrus_sasl
2297 authenticators configured (plain, login, cram-md5):
2299 On startup auth_cyrus_sasl_init() gets called for each of these.
2300 This means three calls to sasl_listmech() without a specified mech_list.
2301 => SASL tests which mechs of all available mechs actually work
2302 => three warnings about OTP not working
2303 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2305 With the patch, sasl_listmech() also gets called three times. But now
2306 SASL's mech_list option is set to the server_mech specified in the the
2307 authenticator. Or in other words, the answer from sasl_listmech()
2308 gets limited to just the mech you're testing for (which is different
2310 => the return list contains just 'plain' or 'login', 'cram-md5' or
2311 nothing depending on the value of ob->server_mech.
2313 I've just tested the patch: Authentication still works fine,
2314 unavailable mechs specified in the exim configuration are still
2315 caught, and the auth.log warnings about OTP are gone."
2317 PH/25 When debugging is enabled, the contents of the command line are added
2318 to the debugging output, even when log_selector=+arguments is not
2321 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2322 answer is "GNU", and only if the return is "GNU/something" is the answer
2325 PH/27 $acl_verify_message is now set immediately after the failure of a
2326 verification in an ACL, and so is available in subsequent modifiers. In
2327 particular, the message can be preserved by coding like this:
2329 warn !verify = sender
2330 set acl_m0 = $acl_verify_message
2332 Previously, $acl_verify_message was set only while expanding "message"
2333 and "log_message" when a very denied access.
2335 PH/28 Modified OS/os.c-Linux with
2337 -#ifndef OS_LOAD_AVERAGE
2338 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2340 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2341 nomenclature these days.)
2343 PH/29 Installed patch from the Sieve maintainer that adds the options
2344 sieve_useraddress and sieve_subaddress to the redirect router.
2346 PH/30 In these circumstances:
2347 . Two addresses routed to the same list of hosts;
2348 . First host does not offer TLS;
2349 . First host accepts first address;
2350 . First host gives temporary error to second address;
2351 . Second host offers TLS and a TLS session is established;
2352 . Second host accepts second address.
2353 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2354 and peerdn, if requested) that were in fact used only for the second
2357 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2358 attention to any local part prefix or suffix that was matched by the
2359 router that accepted the address. It now behaves in the same way as it
2360 does for delivery: the affixes are removed from the local part unless
2361 rcpt_include_affixes is set on the transport.
2363 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2364 timeout during the DATA phase of an incoming message.
2366 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2367 applied a patch sanctioned by the Sieve maintainer.
2369 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2370 the uid or gid is negative. A case of a negative gid caused this to be
2371 noticed. The fix allows for either to be negative.
2373 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2374 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2377 PH/36 The change PH/12 above was broken. Fixed it.
2379 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2380 the grounds that routing the same address twice would always produce the
2381 same answer. This might have been true once, but it is certainly no
2382 longer true now. Routing a child address may depend on the previous
2383 routing that produced that child. Some complicated redirection strategies
2384 went wrong when messages had multiple recipients, and made Exim's
2385 behaviour dependent on the order in which the addresses were given.
2387 I have moved the duplicate checking until after the routing is complete.
2388 Exim scans the addresses that are assigned to local and remote
2389 transports, and removes any duplicates. This means that more work will be
2390 done, as duplicates will always all be routed, but duplicates are
2391 presumably rare, so I don't expect this is of any significance.
2393 For deliveries to pipes, files, and autoreplies, the duplicate checking
2394 still happens during the routing process, since they are not going to be
2397 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2398 It corrects a timeout issue with spamd. This is Ian's comment: "The
2399 background is that sometimes spamd either never reads data from a
2400 connection it has accepted, or it never writes response data. The exiscan
2401 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2402 blindly assumes that writes won't block so it may never time out."
2404 PH/39 Allow G after quota size as well as K and M.
2406 PH/40 The value set for $authenticated_id in an authenticator may not contain
2407 binary zeroes or newlines because the value is written to log lines and
2408 to spool files. There was no check on this. Now the value is run through
2409 the string_printing() function so that such characters are converted to
2410 printable escape sequences.
2412 PH/41 $message_linecount is a new variable that contains the total number of
2413 lines in the message. Compare $body_linecount, which is the count for the
2416 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2417 bounce and delay warning messages, except in certain special cases, which
2420 (a) An SMTP error message from a remote host;
2421 (b) A message specified in a :fail: redirection;
2422 (c) A message specified in a "fail" command in a system filter;
2423 (d) A message specified in a FAIL return from the queryprogram router;
2424 (e) A message specified by the cannot_route_message router option.
2426 In these cases only, Exim does include the error details in bounce and
2427 warning messages. There are also a few cases where bland messages such
2428 as "unrouteable address" or "local delivery error" are given.
2430 PH/43 $value is now also set for the "else" part of a ${run expansion.
2432 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2433 being worked on, but at least Exim now implements the latest version to
2436 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2437 process to complete was treated as a delivery failure, a timeout while
2438 writing the message to the pipe was logged, but erroneously treated as a
2439 successful delivery. Such timeouts include transport filter timeouts. For
2440 consistency with the overall process timeout, these timeouts are now
2441 treated as errors, giving rise to delivery failures by default. However,
2442 there is now a new Boolean option for the pipe transport called
2443 timeout_defer, which, if set TRUE, converts the failures into defers for
2444 both kinds of timeout. A transport filter timeout is now identified in
2447 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2448 systems where "make" and "gmake" are different, calling "gmake" at top
2449 level broke things. I've arranged for the value of $(MAKE) to be passed
2450 from the Makefile to this script so that it can call the same version of
2454 A note about Exim versions 4.44 and 4.50
2455 ----------------------------------------
2457 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2458 changes of various kinds. As a consequence, a big documentation update was
2459 needed. This delayed the release for rather longer than seemed good, especially
2460 in the light of a couple of (minor) security issues. Therefore, the changes
2461 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2462 release. So 4.44 and 4.50 are in effect two different branches that both start
2465 I have left the 4.50 change log unchanged; it contains all the changes since
2466 4.43. The change log for 4.44 is below; many of its items are identical to
2467 those for 4.50. This seems to be the most sensible way to preserve the
2468 historical information.
2474 1. Minor wording change to the doc/README.SIEVE file.
2476 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2477 computation of the current number of files was incorrect.
2479 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2480 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2483 4. Give more explanation in the error message when the command for a transport
2484 filter fails to execute.
2486 5. There are several places where Exim runs a non-Exim command in a
2487 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2488 was being done only for the command run by the queryprogram router. It is
2489 now done for all such subprocesses. The other cases are: ${run, transport
2490 filters, and the commands run by the lmtp and pipe transports.
2492 6. Added CONFIGURE_GROUP build-time option.
2494 7. Some older OS have a limit of 256 on the maximum number of file
2495 descriptors. Exim was using setrlimit() to set 1000 as a large value
2496 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2497 systems. I've change it so that if it can't get 1000, it tries for 256.
2499 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2500 was an oversight, and furthermore, ever since the addition of extra
2501 controls (e.g. 4.43/32), the checks on when to allow different forms of
2502 "control" were broken. There should now be diagnostics for all cases when a
2503 control that does not make sense is encountered.
2505 9. Added the /retain_sender option to "control=submission".
2507 10. $recipients is now available in the predata ACL (oversight).
2509 11. Tidy the search cache before the fork to do a delivery from a message
2510 received from the command line. Otherwise the child will trigger a lookup
2511 failure and thereby defer the delivery if it tries to use (for example) a
2512 cached ldap connection that the parent has called unbind on.
2514 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2515 of $address_data from the recipient verification was clobbered by the
2516 sender verification.
2518 13. The value of address_data from a sender verification is now available in
2519 $sender_address_data in subsequent conditions in the ACL statement.
2521 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2523 15. Added a new option "connect=<time>" to callout options, to set a different
2526 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2527 was its contents. (It was OK if the option was not defined at all.)
2529 17. A "Completed" log line is now written for messages that are removed from
2530 the spool by the -Mrm option.
2532 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2533 information about exactly what failed.
2535 19. Added -dd to debug only the daemon process.
2537 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2538 handles timeouts, both on the server side and network timeouts. Renamed the
2539 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2541 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2542 It is now set to "smtps".
2544 22. $host_address is now set to the target address during the checking of
2545 ignore_target_hosts.
2547 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2548 being passed; this would have caused $sender_host_name to have been used if
2549 matching the list had actually called for a host name (not very likely,
2550 since this list is usually IP addresses). A host name is now passed as
2553 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2554 code that specifies a non-restarting handler (typically sigaction() in
2555 modern systems) in an attempt to fix a rare and obscure crash bug.
2557 25. Narrowed the window for a race in the daemon that could cause it to ignore
2558 SIGCHLD signals. This is not a major problem, because they are used only to
2559 wake it up if nothing else does.
2561 26. A malformed maildirsize file could cause Exim to calculate negative values
2562 for the mailbox size or file count. Odd effects could occur as a result.
2563 The maildirsize information is now recalculated if the size or filecount
2566 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2567 support for a long time. Removed HAVE_SYS_VFS_H.
2569 28. Installed the latest version of exipick from John Jetmore.
2571 29. In an address list, if the pattern was not a regular expression, an empty
2572 subject address (from a bounce message) matched only if the pattern was an
2573 empty string. Non-empty patterns were not even tested. This was the wrong
2574 because it is perfectly reasonable to use an empty address as part of a
2575 database query. An empty address is now tested by patterns that are
2576 lookups. However, all the other forms of pattern expect the subject to
2577 contain a local part and a domain, and therefore, for them, an empty
2578 address still always fails if the pattern is not itself empty.
2580 30. Exim went into a mad DNS loop when attempting to do a callout where the
2581 host was specified on an smtp transport, and looking it up yielded more
2582 than one IP address.
2584 31. Re-factored the code for checking spool and log partition space into a
2585 function that finds that data and another that does the check. The former
2586 is then used to implement four new variables: $spool_space, $log_space,
2587 $spool_inodes, and $log_inodes.
2589 32. The RFC2047 encoding function was originally intended for short strings
2590 such as real names; it was not keeping to the 75-character limit for
2591 encoded words that the RFC imposes. It now respects the limit, and
2592 generates multiple encoded words if necessary. To be on the safe side, I
2593 have increased the buffer size for the ${rfc2047: expansion operator from
2596 33. It is now permitted to omit both strings after an "if" condition; if the
2597 condition is true, the result is "true". As before, when the second string
2598 is omitted, a false condition yields an empty string. This makes it less
2599 cumbersome to write custom ACL and router conditions.
2601 34. Failure to deliver a bounce message always caused it to be frozen, even if
2602 there was an errors_to setting on the router. The errors_to setting is now
2605 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2606 canonical form (fully expanded) before being placed in
2607 $sender_host_address.
2609 36. The table in the code that translates DNS record types into text (T_A to
2610 "A" for instance) was missing entries for NS and CNAME. It is just possible
2611 that this could have caused confusion if both these types were looked up
2612 for the same domain, because the text type is used as part of Exim's
2613 per-process caching. But the chance of anyone hitting this buglet seems
2616 37. The dnsdb lookup has been extended in a number of ways.
2618 (1) There is a new type, "zns", which walks up the domain tree until it
2619 finds some nameserver records. It should be used with care.
2621 (2) There is a new type, "mxh", which is like "mx" except that it returns
2622 just the host names, not the priorities.
2624 (3) It is now possible to give a list of domains (or IP addresses) to be
2625 looked up. The behaviour when one of the lookups defers can be
2626 controlled by a keyword.
2628 (4) It is now possible to specify the separator character for use when
2629 multiple records are returned.
2631 38. The dnslists ACL condition has been extended: it is now possible to supply
2632 a list of IP addresses and/or domains to be looked up in a particular DNS
2635 39. Added log_selector=+queue_time_overall.
2637 40. When running the queue in the test harness, wait just a tad after forking a
2638 delivery process, to get repeatability of debugging output.
2640 41. Include certificate and key file names in error message when GnuTLS fails
2641 to set them up, because the GnuTLS error message doesn't include the name
2642 of the failing file when there is a problem reading it.
2644 42. Allow both -bf and -bF in the same test run.
2646 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2648 44. The "Exiscan patch" is now merged into the mainline Exim source.
2650 45. Sometimes the final signoff response after QUIT could fail to get
2651 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2652 < 0 before doing a fflush(). This bug looks as though it goes back to the
2653 introduction of TLS in release 3.20, but "sometimes" must have been rare
2654 because the tests only now provoked it.
2656 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2657 (this can affect the format of dates).
2659 47. exim_tidydb, when checking for the continued existence of a message for
2660 which it has found a message-specific retry record, was not finding
2661 messages that were in split spool directories. Consequently, it was
2662 deleting retry records that should have stayed in existence.
2664 48. Steve fixed some bugs in eximstats.
2666 49. The SPA authentication driver was not abandoning authentication and moving
2667 on to the next authenticator when an expansion was forced to fail,
2668 contradicting the general specification for all authenticators. Instead it
2669 was generating a temporary error. It now behaves as specified.
2671 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2672 (the order specifies the preference for clients). The order is now AES256,
2673 AES128, 3DES, ARCFOUR128.
2675 51. Small patch to Sieve code - explicitly set From: when generating an
2678 52. Exim crashed if a remote delivery caused a very long error message to be
2679 recorded - for instance if somebody sent an entire SpamAssassin report back
2680 as a large number of 550 error lines. This bug was coincidentally fixed by
2681 increasing the size of one of Exim's internal buffers (big_buffer) that
2682 happened as part of the Exiscan merge. However, to be on the safe side, I
2683 have made the code more robust (and fixed the comments that describe what
2686 53. Now that there can be additional text after "Completed" in log lines (if
2687 the queue_time_overall log selector is set), a one-byte patch to exigrep
2688 was needed to allow it to recognize "Completed" as not the last thing in
2691 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2692 patch that reportedly fixes this has been added. I am not expert enough to
2693 create a test for it. This is what the patch creator wrote:
2695 "I found a little strange behaviour of ldap code when working with
2696 Windows 2003 AD Domain, where users was placed in more than one
2697 Organization Units. When I tried to give exim partial DN, the exit code
2698 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2699 But simultaneously result of request was absolutely normal ldap result,
2700 so I produce this patch..."
2702 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2703 REFERENCE, so I have modified the code to exclude the patch when that macro
2706 55. Some experimental protocols are using DNS PTR records for new purposes. The
2707 keys for these records are domain names, not reversed IP addresses. The
2708 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2709 leaves it alone. Component reversal etc. now happens only for IP addresses.
2712 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2714 57. Double the size of the debug message buffer (to 2048) so that more of very
2715 long debug lines gets shown.
2717 58. The exicyclog utility now does better if the number of log files to keep
2718 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2720 59. Two changes related to the smtp_active_hostname option:
2722 (1) $smtp_active_hostname is now available as a variable.
2723 (2) The default for smtp_banner uses $smtp_active_hostname instead
2724 of $primary_hostname.
2726 60. The host_aton() function is supposed to be passed a string that is known
2727 to be a valid IP address. However, in the case of IPv6 addresses, it was
2728 not checking this. This is a hostage to fortune. Exim now panics and dies
2729 if the condition is not met. A case was found where this could be provoked
2730 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2731 components; fortuitously, this particular loophole had already been fixed
2732 by change 4.50/55 above.
2734 If there are any other similar loopholes, the new check in host_aton()
2735 itself should stop them being exploited. The report I received stated that
2736 data on the command line could provoke the exploit when Exim was running as
2737 exim, but did not say which command line option was involved. All I could
2738 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2739 running as the user.
2742 61. There was a buffer overflow vulnerability in the SPA authentication code
2743 (which came originally from the Samba project). I have added a test to the
2744 spa_base64_to_bits() function which I hope fixes it.
2747 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2748 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2750 63. The daemon start-up calls getloadavg() while still root for those OS that
2751 need the first call to be done as root, but it missed one case: when
2752 deliver_queue_load_max is set with deliver_drop_privilege. This is
2753 necessary for the benefit of the queue runner, because there is no re-exec
2754 when deliver_drop_privilege is set.
2756 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2757 This has been fixed.
2759 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2760 was in use, was not putting the data itself into the right store pool;
2761 consequently, it could be overwritten for a subsequent message in the same
2762 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2765 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2767 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2768 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2769 the function were treating the return as a boolean value, which happened to
2770 work because 0=false and not-0=true, but is not correct code.
2772 68. The host_aton() function was not handling scoped IPv6 addresses (those
2773 with, for example, "%eth0" on the end) correctly.
2775 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2776 negated items (that is, ~something) in unsigned ints. Some compilers
2777 apparently mutter when there is no cast.
2779 70. If an address verification called from an ACL failed, and did not produce a
2780 user-specific message (i.e. there was only a "system" message), nothing was
2781 put in $acl_verify_message. In this situation, it now puts the system
2784 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2785 session; change 4.31/43 added the unwanted input to the log line - except
2786 that it did not do this in the start of session case. It now does.
2788 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2789 This is wrong and can cause the other end to generate a synchronization
2790 error if it is another Exim or anything else that does the synchronization
2791 check. A QUIT command is no longer sent after a timeout.
2793 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2794 during host lookups.
2796 74. The defer_ok option of callout verification was not working if it was used
2797 when verifying addresses in header lines, that is, for this case:
2799 verify = header_sender/callout=defer_ok
2801 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2802 those file descriptors could be used for SMTP connections. If anything
2803 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2804 could be sent to the SMTP client, causing chaos. The daemon now opens
2805 stdin, stdout, and stderr to /dev/null when it puts itself into the
2808 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2809 log by default. The user can override this with suitable Perl magic.
2811 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2812 the log message when discard triggers, was not working for the DATA ACL or
2813 for the non-SMTP ACL.
2815 78. Error message wording change in sieve.c.
2817 79. If smtp_accept_max_per_host was set, the number of connections could be
2818 restricted to fewer than expected, because the daemon was trying to set up
2819 a new connection before checking whether the processes handling previous
2820 connections had finished. The check for completed processes is now done
2821 earlier. On busy systems, this bug wouldn't be noticed because something
2822 else would have woken the daemon, and it would have reaped the completed
2825 80. If a message was submitted locally by a user whose login name contained one
2826 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2827 It caused a spool format error. I have fixed the spool reading code. A
2828 related problem was that the "from" clause in the Received: line became
2829 illegal because of the space(s). It is now covered by ${quote_local_part.
2831 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2834 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2836 83. Patch from Sieve maintainer to fix unterminated string problem in
2837 "vacation" handling.
2839 84. Some minor changes to the Linux configuration files to help with other
2840 OS variants using glibc.
2842 85. One more patch for Sieve to update vacation handling to latest spec.
2845 ----------------------------------------------------
2846 See the note above about the 4.44 and 4.50 releases.
2847 ----------------------------------------------------
2853 1. Change 4.43/35 introduced a bug that caused file counts to be
2854 incorrectly computed when quota_filecount was set in an appendfile
2857 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2858 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2861 3. Give more explanation in the error message when the command for a transport
2862 filter fails to execute.
2864 4. There are several places where Exim runs a non-Exim command in a
2865 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2866 was being done only for the command run by the queryprogram router. It is
2867 now done for all such subprocesses. The other cases are: ${run, transport
2868 filters, and the commands run by the lmtp and pipe transports.
2870 5. Some older OS have a limit of 256 on the maximum number of file
2871 descriptors. Exim was using setrlimit() to set 1000 as a large value
2872 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2873 systems. I've change it so that if it can't get 1000, it tries for 256.
2875 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2876 was an oversight, and furthermore, ever since the addition of extra
2877 controls (e.g. 4.43/32), the checks on when to allow different forms of
2878 "control" were broken. There should now be diagnostics for all cases when a
2879 control that does not make sense is encountered.
2881 7. $recipients is now available in the predata ACL (oversight).
2883 8. Tidy the search cache before the fork to do a delivery from a message
2884 received from the command line. Otherwise the child will trigger a lookup
2885 failure and thereby defer the delivery if it tries to use (for example) a
2886 cached ldap connection that the parent has called unbind on.
2888 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2889 of $address_data from the recipient verification was clobbered by the
2890 sender verification.
2892 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2893 was its contents. (It was OK if the option was not defined at all.)
2895 11. A "Completed" log line is now written for messages that are removed from
2896 the spool by the -Mrm option.
2898 12. $host_address is now set to the target address during the checking of
2899 ignore_target_hosts.
2901 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2902 being passed; this would have caused $sender_host_name to have been used if
2903 matching the list had actually called for a host name (not very likely,
2904 since this list is usually IP addresses). A host name is now passed as
2907 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2908 code that specifies a non-restarting handler (typically sigaction() in
2909 modern systems) in an attempt to fix a rare and obscure crash bug.
2911 15. Narrowed the window for a race in the daemon that could cause it to ignore
2912 SIGCHLD signals. This is not a major problem, because they are used only to
2913 wake it up if nothing else does.
2915 16. A malformed maildirsize file could cause Exim to calculate negative values
2916 for the mailbox size or file count. Odd effects could occur as a result.
2917 The maildirsize information is now recalculated if the size or filecount
2920 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2921 support for a long time. Removed HAVE_SYS_VFS_H.
2923 18. Updated exipick to current release from John Jetmore.
2925 19. Allow an empty sender to be matched against a lookup in an address list.
2926 Previously the only cases considered were a regular expression, or an
2929 20. Exim went into a mad DNS lookup loop when doing a callout where the
2930 host was specified on the transport, if the DNS lookup yielded more than
2933 21. The RFC2047 encoding function was originally intended for short strings
2934 such as real names; it was not keeping to the 75-character limit for
2935 encoded words that the RFC imposes. It now respects the limit, and
2936 generates multiple encoded words if necessary. To be on the safe side, I
2937 have increased the buffer size for the ${rfc2047: expansion operator from
2940 22. Failure to deliver a bounce message always caused it to be frozen, even if
2941 there was an errors_to setting on the router. The errors_to setting is now
2944 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2945 canonical form (fully expanded) before being placed in
2946 $sender_host_address.
2948 24. Updated eximstats to version 1.33
2950 25. Include certificate and key file names in error message when GnuTLS fails
2951 to set them up, because the GnuTLS error message doesn't include the name
2952 of the failing file when there is a problem reading it.
2954 26. Expand error message when OpenSSL has problems setting up cert/key files.
2957 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2958 (this can affect the format of dates).
2960 28. exim_tidydb, when checking for the continued existence of a message for
2961 which it has found a message-specific retry record, was not finding
2962 messages that were in split spool directories. Consequently, it was
2963 deleting retry records that should have stayed in existence.
2965 29. eximstats updated to version 1.35
2966 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2967 1.35 - bugfix such that pie charts by volume are generated correctly
2969 30. The SPA authentication driver was not abandoning authentication and moving
2970 on to the next authenticator when an expansion was forced to fail,
2971 contradicting the general specification for all authenticators. Instead it
2972 was generating a temporary error. It now behaves as specified.
2974 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2975 (the order specifies the preference for clients). The order is now AES256,
2976 AES128, 3DES, ARCFOUR128.
2978 31. Small patch to Sieve code - explicitly set From: when generating an
2981 32. Exim crashed if a remote delivery caused a very long error message to be
2982 recorded - for instance if somebody sent an entire SpamAssassin report back
2983 as a large number of 550 error lines. This bug was coincidentally fixed by
2984 increasing the size of one of Exim's internal buffers (big_buffer) that
2985 happened as part of the Exiscan merge. However, to be on the safe side, I
2986 have made the code more robust (and fixed the comments that describe what
2989 33. Some experimental protocols are using DNS PTR records for new purposes. The
2990 keys for these records are domain names, not reversed IP addresses. The
2991 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2992 leaves it alone. Component reversal etc. now happens only for IP addresses.
2995 34. The host_aton() function is supposed to be passed a string that is known
2996 to be a valid IP address. However, in the case of IPv6 addresses, it was
2997 not checking this. This is a hostage to fortune. Exim now panics and dies
2998 if the condition is not met. A case was found where this could be provoked
2999 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3000 components; fortuitously, this particular loophole had already been fixed
3001 by change 4.50/55 or 4.44/33 above.
3003 If there are any other similar loopholes, the new check in host_aton()
3004 itself should stop them being exploited. The report I received stated that
3005 data on the command line could provoke the exploit when Exim was running as
3006 exim, but did not say which command line option was involved. All I could
3007 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3008 running as the user.
3011 35. There was a buffer overflow vulnerability in the SPA authentication code
3012 (which came originally from the Samba project). I have added a test to the
3013 spa_base64_to_bits() function which I hope fixes it.
3016 36. The daemon start-up calls getloadavg() while still root for those OS that
3017 need the first call to be done as root, but it missed one case: when
3018 deliver_queue_load_max is set with deliver_drop_privilege. This is
3019 necessary for the benefit of the queue runner, because there is no re-exec
3020 when deliver_drop_privilege is set.
3022 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3023 was in use, was not putting the data itself into the right store pool;
3024 consequently, it could be overwritten for a subsequent message in the same
3025 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3028 38. Sometimes the final signoff response after QUIT could fail to get
3029 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3030 < 0 before doing a fflush(). This bug looks as though it goes back to the
3031 introduction of TLS in release 3.20, but "sometimes" must have been rare
3032 because the tests only now provoked it.
3038 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3039 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3040 Now it returns an integer. A number of calls were still expecting a T/F
3041 return. Fortuitously, in all cases, the tests worked in OK situations,
3042 which is the norm. However, things would have gone wrong on any write
3043 failures on the smtp file descriptor. This function is used when sending
3044 messages over SMTP and also when doing verify callouts.
3046 2. When Exim is called to do synchronous delivery of a locally submitted
3047 message (the -odf or -odi options), it no longer closes stderr before doing
3050 3. Implemented the mua_wrapper option.
3052 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3054 5. Implemented the functions header_remove(), header_testname(),
3055 header_add_at_position(), and receive_remove_recipient(), and exported them
3058 6. If an ACL "warn" statement specified the addition of headers, Exim already
3059 inserted X-ACL-Warn: at the start if there was no header name. However, it
3060 was not making this test for the second and subsequent header lines if
3061 there were newlines in the string. This meant that an invalid header could
3062 be inserted if Exim was badly configured.
3064 7. Allow an ACL "warn" statement to add header lines at the start or after all
3065 the Received: headers, as well as at the end.
3067 8. Added the rcpt_4xx retry error code.
3069 9. Added postmaster_mailfrom=xxx to callout verification option.
3071 10. Added mailfrom=xxxx to the callout verification option, for verify=
3074 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3075 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3077 12. Inserted some casts to stop certain compilers warning when using pointer
3078 differences as field lengths or precisions in printf-type calls (mostly
3079 affecting debugging statements).
3081 13. Added optional readline() support for -be (dynamically loaded).
3083 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3084 same clock tick as a message's arrival, so that its received time was the
3085 same as the "first fail" time on the retry record, and that message
3086 remained on the queue past the ultimate address timeout, every queue runner
3087 would try a delivery (because it was past the ultimate address timeout) but
3088 after another failure, the ultimate address timeout, which should have then
3089 bounced the address, did not kick in. This was a "< instead of <=" error;
3090 in most cases the first failure would have been in the next clock tick
3091 after the received time, and all would be well.
3093 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3094 being recognized when the domain list was tested by the match_domain
3095 condition in an expansion string.
3097 16. Added the ${str2b64: operator.
3099 17. Exim was always calling setrlimit() to set a large limit for the number of
3100 processes, without checking whether the existing limit was already
3101 adequate. (It did check for the limit on file descriptors.) Furthermore,
3102 errors from getrlimit() and setrlimit() were being ignored. Now they are
3103 logged to the main and panic logs, but Exim does carry on, to try to do its
3104 job under whatever limits there are.
3106 18. Imported PCRE 5.0.
3108 19. Trivial typo in log message " temporarily refused connection" (the leading
3111 20. If the log selector return_path_on_delivery was set and an address was
3112 redirected to /dev/null, the delivery process crashed because it assumed
3113 that a return path would always be set for a "successful" delivery. In this
3114 case, the whole delivery is bypassed as an optimization, and therefore no
3117 21. Internal re-arrangement: the function for sending a challenge and reading
3118 a response while authentication was assuming a zero-terminated challenge
3119 string. It's now changed to take a pointer and a length, to allow for
3120 binary data in such strings.
3122 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3124 23. Exim was not respecting finduser_retries when seeking the login of the
3125 uid under which it was called; it was always trying 10 times. (The default
3126 setting of finduser_retries is zero.) Also, it was sleeping after the final
3127 failure, which is pointless.
3129 24. Implemented tls_on_connect_ports.
3131 25. Implemented acl_smtp_predata.
3133 26. If the domain in control=submission is set empty, Exim assumes that the
3134 authenticated id is a complete email address when it generates From: or
3135 Sender: header lines.
3137 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3138 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3139 chown and chgrp in /bin and hostname in /usr/bin.
3141 28. Exim was keeping the "process log" file open after each use, just as it
3142 does for the main log. This opens the possibility of it remaining open for
3143 long periods when the USR1 signal hits a daemon. Occasional processlog
3144 errors were reported, that could have been caused by this. Anyway, it seems
3145 much more sensible not to leave this file open at all, so that is what now
3148 29. The long-running daemon process does not normally write to the log once it
3149 has entered its main loop, and it closes the log before doing so. This is
3150 so that log files can straightforwardly be renamed and moved. However,
3151 there are a couple of unusual error situations where the daemon does write
3152 log entries, and I had neglected to close the log afterwards.
3154 30. The text of an SMTP error response that was received during a remote
3155 delivery was being truncated at 512 bytes. This is too short for some of
3156 the long messages that one sometimes sees. I've increased the limit to
3159 31. It is now possible to make retry rules that apply only when a message has a
3160 specific sender, in particular, an empty sender.
3162 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3163 it possible to be selective about when SMTP synchronization is enforced.
3165 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3167 32. Implemented hosts_connection_nolog.
3169 33. Added an ACL for QUIT.
3171 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3174 35. Added mailbox_size and mailbox_filecount to appendfile.
3176 36. Added control = no_multiline_responses to ACLs.
3178 37. There was a bug in the logic of the code that waits for the clock to tick
3179 in the case where the clock went backwards by a substantial amount such
3180 that the microsecond fraction of "now" was more than the microsecond
3181 fraction of "then" (but the whole seconds number was less).
3183 38. Added support for the libradius Radius client library this is found on
3184 FreeBSD (previously only the radiusclient library was supported).
3190 1. When certain lookups returned multiple values in the form name=value, the
3191 quoting of the values was not always being done properly. Specifically:
3192 (a) If the value started with a double quote, but contained no whitespace,
3194 (b) If the value contained whitespace other than a space character (i.e.
3195 tabs or newlines or carriage returns) it was not quoted.
3196 This fix has been applied to the mysql and pgsql lookups by writing a
3197 separate quoting function and calling it from the lookup code. The fix
3198 should probably also be applied to nisplus, ibase and oracle lookups, but
3199 since I cannot test any of those, I have not disturbed their existing code.
3201 2. A hit in the callout cache for a specific address caused a log line with no
3202 reason for rejecting RCPT. Now it says "Previous (cached) callout
3203 verification failure".
3205 3. There was an off-by-one bug in the queryprogram router. An over-long
3206 return line was truncated at 256 instead of 255 characters, thereby
3207 overflowing its buffer with the terminating zero. As well as fixing this, I
3208 have increased the buffer size to 1024 (and made a note to document this).
3210 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3211 when Exim is waiting for an SMTP response from a remote server, Exim
3212 restarts its select() call on the socket, thereby resetting its timeout.
3213 This is not a problem when such interrupts are rare. Somebody set up a cron
3214 job to run exiwhat every 2 minutes, which is less than the normal select()
3215 timeout (5 or 10 minutes). This meant that the select() timeout never
3216 kicked in because it was always reset. I have fixed this by comparing the
3217 time when an interrupt arrives with the time at the start of the first call
3218 to select(). If more time than the timeout has elapsed, the interrupt is
3219 treated as a timeout.
3221 5. Some internal re-factoring in preparation for the addition of Sieve
3222 extensions (by MH). In particular, the "personal" test is moved to a
3223 separate function, and given an option for scanning Cc: and Bcc: (which is
3224 not set for Exim filters).
3226 6. When Exim created an email address using the login of the caller as the
3227 local part (e.g. when creating a From: or Sender: header line), it was not
3228 quoting the local part when it contained special characters such as @.
3230 7. Installed new OpenBSD configuration files.
3232 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3233 try to make them clearer.
3235 9. Callout options, other than the timeout value, were being ignored when
3236 verifying sender addresses in header lines. For example, when using
3238 verify = header_sender/callout=no_cache
3240 the cache was (incorrectly) being used.
3242 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3243 only the Cygwin environment.
3245 11. When return_path_on_delivery was set as a log selector, if different remote
3246 addresses in the same message used different return paths and parallel
3247 remote delivery occurred, the wrong values would sometimes be logged.
3248 (Whenever a remote delivery process finished, the return path value from
3249 the most recently started remote delivery process was logged.)
3251 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3252 lines when AUTH and/or TLS are in use. This is the "received protocol"
3253 field. Exim used to use "asmtp" for authenticated SMTP, without any
3254 indication (in the protocol name) for TLS use. Now it follows the RFC and
3255 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3256 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3257 names appear in log lines as well as in Received: header lines.
3259 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3260 extensions, and comparison tests, and to fix some bugs.
3262 14. Changes to the "personal" filter test:
3264 (1) The test was buggy in that it was just doing the equivalent of
3265 "contains" tests on header lines. For example, if a user's address was
3266 anne@some.where, the "personal" test would incorrectly be true for
3268 To: susanne@some.where
3270 This test is now done by extracting each address from the header in turn,
3271 and checking the entire address. Other tests that are part of "personal"
3272 are now done using regular expressions (for example, to check local parts
3273 of addresses in From: header lines).
3275 (2) The list of non-personal local parts in From: addresses has been
3276 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3277 taken from the Sieve specification recommendations.
3279 (3) If the message contains any header line starting with "List-" it is
3280 treated as non-personal.
3282 (4) The test for "circular" in the Subject: header line has been removed
3283 because it now seems ill-conceived.
3285 15. Minor typos in src/EDITME comments corrected.
3287 16. Installed latest exipick from John Jetmore.
3289 17. If headers_add on a router specified a text string that was too long for
3290 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3291 of string_sprintf() is now avoided.
3293 18. $message_body_size was not set (it was always zero) when running the DATA
3294 ACL and the local_scan() function.
3296 19. For the "mail" command in an Exim filter, no default was being set for
3297 the once_repeat time, causing a random time value to be used if "once" was
3298 specified. (If the value happened to be <= 0, no repeat happened.) The
3299 default is now 0s, meaning "never repeat". The "vacation" command was OK
3300 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3301 (I found it when inspecting the code).
3303 20. There is now an overall timeout for performing a callout verification. It
3304 defaults to 4 times the callout timeout, which applies to individual SMTP
3305 commands during the callout. The overall timeout applies when there is more
3306 than one host that can be tried. The timeout is checked before trying the
3307 next host. This prevents very long delays if there are a large number of
3308 hosts and all are timing out (e.g. when the network connections are timing
3309 out). The value of the overall timeout can be changed by specifying an
3310 additional sub-option for "callout", called "maxwait". For example:
3312 verify = sender/callout=5s,maxwait=20s
3314 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3315 to the end before writing, but this should make it even safer).
3317 22. Exim was forgetting that it had advertised PIPELINING for the second and
3318 subsequent messages on an SMTP connection. It was also not resetting its
3319 memory on STARTTLS and an internal HELO.
3321 23. When Exim logs an SMTP synchronization error within a session, it now
3322 records whether PIPELINING has been advertised or not.
3324 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3325 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3326 rather than long int.
3328 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3330 26. Added the never_mail option to autoreply.
3336 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3337 crash if the getsockname() call failed; this can happen if a connection is
3338 closed very soon after it is established. The problem was simply in the
3339 order in which certain operations were done, causing Exim to try to write
3340 to the SMTP stream before it had set up the file descriptor. The bug has
3341 been fixed by making things happen in the correct order.
3347 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3348 before the connection was closed, thus losing the rejection response.
3350 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3351 some early Solaris releases, but causes trouble in current releases where
3352 socklen_t is defined.
3354 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3357 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3360 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3361 that releasing the top of it at the end releases what was used for sub-
3362 expansions (unless the block got too big). However, discard this block if
3363 the first thing is a variable or header, so that we can use its block when
3364 it is dynamic (useful for very large $message_headers, for example).
3366 6. Lookups now cache *every* query, not just the most recent. A new, separate
3367 store pool is used for this. It can be recovered when all lookup caches are
3368 flushed. Lookups now release memory at the end of their result strings.
3369 This has involved some general refactoring of the lookup sources.
3371 7. Some code has been added to the store_xxx() functions to reduce the amount
3372 of flapping under certain conditions.
3374 8. log_incoming_interface used to affect only the <= reception log lines. Now
3375 it causes the local interface and port to be added to several more SMTP log
3376 lines, for example "SMTP connection from", and rejection lines.
3378 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3380 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3382 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3383 could be overwritten at the end of the current message (or the start of a
3384 new message if it was set in a HELO ACL). The value is now preserved for
3385 the duration of the SMTP connection.
3387 12. If a transport had a headers_rewrite setting, and a matching header line
3388 contained an unqualified address, that address was qualified, even if it
3389 did not match any rewriting rules. The underlying bug was that the values
3390 of the flags that permit the existence of unqualified sender and recipient
3391 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3392 non-local messages, and by -bnq for local messages) were not being
3393 preserved with the message after it was received.
3395 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3396 "next input=" as part of the text comprising the host identity instead of
3397 the correct text. The code was using the same buffer for two different
3398 strings. However, depending on which order the printing function evaluated
3399 its arguments, the bug did not always show up. Under Linux, for example, my
3400 test suite worked just fine.
3402 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3403 doesn't work with some older versions of Perl. It has been changed to "my",
3404 which in any case is probably the better facility to use.
3406 15. A really picky compiler found some instances of statements for creating
3407 error messages that either had too many or two few arguments for the format
3410 16. The size of the buffer for calls to the DNS resolver has been increased
3411 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3412 for addresses that have a lot of PTR records. This alleviates a problem; it
3413 does not fully solve it.
3415 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3416 buffer now truncates the list and logs the incident, which is the same
3417 action as happens when Exim is looking up a host name and its aliases.
3418 Previously in this situation something unpredictable would happen;
3419 sometimes it was "internal error: store_reset failed".
3421 18. If a server dropped the connection unexpectedly when an Exim client was
3422 using GnuTLS and trying to read a response, the client delivery process
3423 crashed while trying to generate an error log message.
3425 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3426 string, for example:
3428 warn message = H1: something\nH2: something
3430 the text was added as a single header line from Exim's point of view
3431 though it ended up OK in the delivered message. However, searching for the
3432 second and subsequent header lines using $h_h2: did not work. This has been
3433 fixed. Similarly, if a system filter added multiple headers in this way,
3434 the routers could not see them.
3436 20. Expanded the error message when iplsearch is called with an invalid key to
3437 suggest using net-iplsearch in a host list.
3439 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3440 ACLs are no longer actually imposed (and a message to that effect is
3443 22. If a "gecos" field in a passwd entry contained escaped characters, in
3444 particular, if it contained a \" sequence, Exim got it wrong when building
3445 a From: or a Sender: header from that name. A second bug also caused
3446 incorrect handling when an unquoted " was present following a character
3447 that needed quoting.
3449 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3450 was not being matched caselessly.
3452 24. Arranged for all hyphens in the exim.8 source to be escaped with
3455 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3456 properly. Recipient callout cache records were still being keyed to include
3457 the sender, even when use_sender was set false. This led to far more
3458 callouts that were necessary. The sender is no longer included in the key
3459 when use_sender is false.
3461 26. Added "control = submission" modifier to ACLs.
3463 27. Added the ${base62d: operator to decode base 62 numbers.
3465 28. dnsdb lookups can now access SRV records.
3467 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3468 the configuration file.
3470 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3471 (-v) mode. This makes the output for a verbose queue run more intelligible.
3473 31. Added a use_postmaster feature to recipient callouts.
3475 32. Added the $body_zerocount variable, containing the number of binary zero
3476 bytes in the message body.
3478 33. The time of last modification of the "new" subdirectory is now used as the
3479 "mailbox time last read" when there is a quota error for a maildir
3482 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3484 35. Added +ignore_unknown as a special item in host lists.
3486 36. Code for decoding IPv6 addresses in host lists is now included, even if
3487 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3488 address was recognized as an IP address, but was then not correctly decoded
3489 into binary, causing unexpected and incorrect effects when compared with
3496 1. Very minor rewording of debugging text in manualroute to say "list of
3497 hosts" instead of "hostlist".
3499 2. If verify=header_syntax was set, and a header line with an unqualified
3500 address (no domain) and a large number of spaces between the end of the
3501 name and the colon was received, the reception process suffered a buffer
3502 overflow, and (when I tested it) crashed. This was caused by some obsolete
3503 code that should have been removed. The fix is to remove it!
3505 3. When running in the test harness, delay a bit after writing a bounce
3506 message to get a bit more predictability in the log output.
3508 4. Added a call to search_tidyup() just before forking a reception process. In
3509 theory, someone could use a lookup in the expansion of smtp_accept_max_
3510 per_host which, without the tidyup, could leave open a database connection.
3512 5. Added the variables $recipient_data and $sender_data which get set from a
3513 lookup success in an ACL "recipients" or "senders" condition, or a router
3514 "senders" option, similar to $domain_data and $local_part_data.
3516 6. Moved the writing of debug_print from before to after the "senders" test
3519 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3520 problems for message scanning, either using a data ACL, or using
3521 local_scan() because the Received: header was not generated till after they
3522 were called (in order to set the time as the time of reception completion).
3523 I have revised the way this works. The header is now generated after the
3524 body is received, but before the ACL or local_scan() are called. After they
3525 are run, the timestamp in the header is updated.
3531 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3532 before starting a queue runner without re-exec. This happened only when
3533 deliver_drop_privilege was set or when the Exim user was set to root. The
3534 effect of the bug was that timeouts during subsequent deliveries caused
3535 crashes instead of being properly handled. The handler is now left at its
3536 default (and expected) setting.
3538 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3539 message, again when deliver_drop_privilege is set or Exim is run as root.
3540 The bug described in (1) was not present in this case, but the tidying up
3541 of the other signals was missing. I have made the two cases consistent.
3543 3. The ignore_target_hosts setting on a manualroute router was being ignored
3544 for hosts that were looked up using the /MX notation.
3546 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3549 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3550 operated on the sender address. After changing the $sender_address to <>
3551 for the sender address verify, Exim was re-instated it as the original
3552 (before rewriting) address, but remembering that it had rewritten it, so it
3553 wasn't rewriting it again. This bug also had the effect of breaking the
3554 sender address verification caching when the sender address was rewritten.
3556 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3557 This has been changed so that if the ip literal address matches
3558 ignore_target_hosts, the router declines.
3560 7. Added expansion conditions match_domain, match_address, and match_local_
3561 part (NOT match_host).
3563 8. The placeholder for the Received: header didn't have a length field set.
3565 9. Added code to Exim itself and to exim_lock to test for a specific race
3566 condition that could lead to file corruption when using MBX delivery. The
3567 issue is with the lockfile that is created in /tmp. If this file is removed
3568 after a process has opened it but before that process has acquired a lock,
3569 there is the potential for a second process to recreate the file and also
3570 acquire a lock. This could lead to two Exim processes writing to the file
3571 at the same time. The added code performs the same test as UW imapd; it
3572 checks after acquiring the lock that its file descriptor still refers to
3573 the same named file.
3575 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3576 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3579 11. Added the smtp_active_hostname option. If used, this will typically be made
3580 to depend on the incoming interface address. Because $interface_address is
3581 not set up until the daemon has forked a reception process, error responses
3582 that can happen earlier (such as "too many connections") no longer contain
3585 12. If an expansion in a condition on a "warn" statement fails because a lookup
3586 defers, the "warn" statement is abandoned, and the next ACL statement is
3587 processed. Previously this caused the whole ACL to be aborted.
3589 13. Added the iplsearch lookup type.
3591 14. Added ident_timeout as a log selector.
3593 15. Added tls_certificate_verified as a log selector.
3595 16. Added a global option tls_require_ciphers (compare the smtp transport
3596 option of the same name). This controls incoming TLS connections.
3598 17. I finally figured out how to make tls_require_ciphers do a similar thing
3599 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3600 before starting the TLS session.
3602 18. Tabs are now shown as \t in -bP output.
3604 19. If the log selector return_path_on_delivery was set, Exim crashed when
3605 bouncing a message because it had too many Received: header lines.
3607 20. If two routers both had headers_remove settings, and the first one included
3608 a superfluous trailing colon, the final name in the first list and the
3609 first name in the second list were incorrectly joined into one item (with a
3610 colon in the middle).
3616 1. Added -C and -D options to the exinext utility, mainly to make it easier
3617 to include in the automated testing, but these could be helpful when
3618 multiple configurations are in use.
3620 2. The exinext utility was not formatting the output nicely when there was
3621 an alternate port involved in the retry record key, nor when there was a
3622 message id as well (for retries that were specific to a specific message
3623 and a specific host). It was also confused by IPv6 addresses, because of
3624 the additional colons they contain. I have fixed the IPv4 problem, and
3625 patched it up to do a reasonable job for IPv6.
3627 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3628 delivery, the log line now contains "pipelined" if PIPELINING was used.
3630 4. An SMTP transport process used to panic and die if the bind() call to set
3631 an explicit outgoing interface failed. This has been changed; it is now
3632 treated in the same way as a connect() failure.
3634 5. A reference to $sender_host_name in the part of a conditional expansion
3635 that was being skipped was still causing a DNS lookup. This no longer
3638 6. The def: expansion condition was not recognizing references to header lines
3639 that used bh_ and bheader_.
3641 7. Added the _cache feature to named lists.
3643 8. The code for checking quota_filecount in the appendfile transport was
3644 allowing one more file than it should have been.
3646 9. For compatibility with Sendmail, the command line option
3654 and sets the incoming protocol and host name (for trusted callers). The
3655 host name and its colon can be omitted when only the protocol is to be set.
3656 Note the Exim already has two private options, -pd and -ps, that refer to
3657 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3658 "s", but I don't think that's a major issue.
3660 10. A number of refactoring changes to the code, none of which should affect
3663 (a) The number of logging options was getting close to filling up the
3664 32-bit word that was used as a bit map. I have split them into two classes:
3665 those that are passed in the argument to log_write(), and those that are
3666 only ever tested independently outside of that function. These are now in
3667 separate 32-bit words, so there is plenty of room for expansion again.
3668 There is no change in the user interface or the logging behaviour.
3670 (b) When building, for example, log lines, the code previously used a
3671 macro that called string_cat() twice, in order to add two strings. This is
3672 not really sufficiently general. Furthermore, there was one instance where
3673 it was actually wrong because one of the argument was used twice, and in
3674 one call a function was used. (As it happened, calling the function twice
3675 did not affect the overall behaviour.) The macro has been replaced by a
3676 function that can join an arbitrary number of extra strings onto a growing
3679 (c) The code for expansion conditions now uses a table and a binary chop
3680 instead of a serial search (which was left over from when there were very
3681 few conditions). Also, it now recognizes conditions like "pam" even when
3682 the relevant support is not compiled in: a suitably worded error message is
3683 given if an attempt is made to use such a condition.
3685 11. Added ${time_interval:xxxxx}.
3687 12. A bug was causing one of the ddress fields not to be passed back correctly
3688 from remote delivery subprocesses. The field in question was not being
3689 subsequently used, so this caused to problems in practice.
3691 13. Added new log selectors queue_time and deliver_time.
3693 14. Might have fixed a bug in maildirsizefile handling that threw up
3694 "unexpected character" debug warnings, and recalculated the data
3695 unnecessarily. In any case, I expanded the warning message to give more
3698 15. Added the message "Restricted characters in address" to the statements in
3699 the default ACL that block characters like @ and % in local parts.
3701 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3702 Three changes have been made:
3704 (a) There was a serious bug; a negative response to MAIL caused the whole
3705 recipient domain to be cached as invalid, thereby blocking all messages
3706 to all local parts at the same domain, from all senders. This bug has
3707 been fixed. The domain is no longer cached after a negative response to
3708 MAIL if the sender used is not empty.
3710 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3713 (c) A new callout option, "use_sender" has been added for people who want
3714 the modified behaviour.
3720 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3723 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3724 indeed breaks things for older releases.
3726 3. Added additional logging to the case where there is a problem reading data
3727 from a filter that is running in a subprocess using a pipe, in order to
3728 try to track down a specific problem.
3730 4. Testing facility fudge: when running in the test harness and attempting
3731 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3732 getting "No route to host". Convert this to a timeout.
3734 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3737 6. Some OS don't have socklen_t but use size_t instead. This affects the
3738 fifth argument of getsockopt() amongst other things. This is now
3739 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3740 can be set for individual OS. I have set it for SunOS5, OSF1, and
3741 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3742 some earlier ones do not.
3744 7. Change 4.30/15 was not doing the test caselessly.
3746 8. The standard form for an IPv6 address literal was being rejected by address
3747 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3748 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3749 this, as well as the form without the "IPv6" on the front (but only when
3750 address literals are enabled, of course).
3752 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3754 10. Exim crashed if a message with an empty sender address specified by -f
3755 encountered a router with an errors_to setting. This could be provoked only
3756 by a command such as
3760 where an empty string was supplied; "<>" did not hit this bug.
3762 11. Installed PCRE release 4.5.
3764 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3765 remained set. It is now erased.
3767 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3768 times from message ids (which are base 36 rather than the normal 62).
3770 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3771 were being counted as actual protocol errors, and logged if the log
3772 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3773 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3774 and DATA following a set of rejected RCPTs do not count as protocol errors.
3775 In other words, Exim assumes they were pipelined, though this may not
3776 actually be the case. Of course, in all cases the client gets an
3777 appropriate error code.
3779 15. If a lookup fails in an ACL condition, a message about the failure may
3780 be available; it is used if testing the ACL cannot continue, because most
3781 such messages specify what the cause of the deferral is. However, some
3782 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3783 that caused an old message to be retained and used if a later statement
3784 caused a defer, replacing the real cause of the deferral.
3786 16. If an IP address had so many PTR records that the DNS lookup buffer
3787 was not large enough to hold them, Exim could crash while trying to process
3788 the truncated data. It now detects and logs this case.
3790 17. Further to 4.21/58, another change has been made: if (and only if) the
3791 first line of a message (the first header line) ends with CRLF, a bare LF
3792 in a subsequent header line has a space inserted after it, so as not to
3793 terminate the header.
3795 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3796 unnecessarily, used atoi() instead of strtol(), and didn't check the
3797 termination when getting file sizes from file names by regex.
3799 19. Completely re-implemented the support for maildirsize files, in the light
3800 of a number of problems with the previous contributed implementation
3801 (4.30/29). In particular:
3803 . If the quota is zero, the maildirsize file is maintained, but no quota is
3806 . If the maildir directory does not exist, it is created before any attempt
3807 to write a maildirsize file.
3809 . The quota value in the file is just a cache; if the quota is changed in
3810 the transport, the new value overrides.
3812 . A regular expression is available for excluding directories from the
3815 20. The autoreply transport checks the characters in options that define the
3816 message's headers; it allows continued headers, but it was checking with
3817 isspace() after an embedded newline instead of explicitly looking for a
3820 21. If all the "regular" hosts to which an address was routed had passed their
3821 expiry times, and had not reached their retry times, the address was
3822 bounced, even if fallback hosts were defined. Now Exim should go on to try
3825 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3826 equivalent code in the SMTP transport. Some hosts send humungous responses
3827 to HELO/EHLO, more than 1024 it seems.
3829 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3830 gives compiler warnings in some environments. I've now done it "properly",
3833 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3834 (because of problems with the built-in one) was declared to return uschar *
3835 instead of char *, causing compiler failure.
3837 25. Fixed a file descriptor leak when processing alias/forward files.
3839 26. Fixed a minor format string issue in dbfn.c.
3841 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3843 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3844 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3847 29. When a sender address is verified, it is cached, to save repeating the test
3848 when there is more than one recipient in a message. However, when the
3849 verification involves a callout, it is possible for different callout
3850 options to be set for different recipients. It is too complicated to keep
3851 track of this in the cache, so now Exim always runs a verification when a
3852 callout is required, relying on the callout cache for the optimization.
3853 The overhead is duplication of the address routing, but this should not be
3856 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3857 to be verified with callout=postmaster, and the main callout worked but the
3858 postmaster check failed, the verification correctly failed. However, if a
3859 subsequent RCPT command asked for sender verification *without* the
3860 postmaster check, incorrect caching caused this verification also to fail,
3863 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3864 it was not caching the DNS options (qualify_single, search_parents) that
3865 were used when the lookup failed. A subsequent lookup with different
3866 options therefore always gave the same answer, though there were cases
3867 where it should not have. (Example: a "domains = !$mx_any" option on a
3868 dnslookup router: the "domains" option is always processed without any
3869 widening, but the router might have qualify_single set.) Now Exim uses the
3870 cached value only when the same options are set.
3872 32. Added John Jetmore's "exipick" utility to the distribution.
3874 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3875 than a timeout (e.g. a certificate is required, and is not provided), an
3876 Exim server now closes the connection immediately. Previously it waited for
3877 the client to close - but if the client is SSL, it seems that they each
3878 wait for each other, leading to a delay before one of them times out.
3880 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3881 maintained 0.8.x compatibility because I don't think many are using it, and
3882 it is clearly obsolete.
3884 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3887 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3888 client certificate was expired. A simple patch fixes this, though I don't
3889 understand the full logic of why the verify callback is called multiple
3892 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3893 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3894 which causes problems with some clients (such as the Certicom SSL Plus
3895 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3896 disables the coutermeasure allowing Eudora to connect."
3898 38. Exim was not checking that a write() to a log file succeeded. This could
3899 lead to Bad Things if a log got too big, in particular if it hit a file
3900 size limit. Exim now panics and dies if it cannot write to a log file, just
3901 as it does if it cannot open a log file.
3903 39. Modified OS/Makefile-Linux so that it now contains
3905 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3907 The two -D definitions ensure that Exim is compiled with large file
3908 support, which makes it possible to handle log files that are bigger than
3911 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3912 instance) a domain was checked against a named list that involved a lookup,
3913 causing $domain_data to be set, then another domain was checked against the
3914 same list, then the first domain was re-checked, the value of $domain_data
3915 after the final check could be wrong. In particular, if the second check
3916 failed, it could be set empty. This bug probably also applied to
3919 41. The strip_trailing_dot option was not being applied to the address given
3920 with the -f command-line option.
3922 42. The code for reading a message's header from the spool was incrementing
3923 $received_count, but never initializing it. This meant that the value was
3924 incorrect (doubled) while delivering a message in the same process in which
3925 it was received. In the most common configuration of Exim, this never
3926 happens - a fresh exec is done - but it can happen when
3927 deliver_drop_privilege is set.
3929 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3930 it now includes up to 150 characters of the unexpected data in the log
3933 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3934 and building data strings. The size of both of these buffers was 10 000
3935 bytes - far larger than anybody would *ever* want, thought I. Needless to
3936 say, somebody hit the limit. I have increased the maximum line length to
3937 20 000 and the maximum data length of concatenated lines to 100 000. I have
3938 also fixed two bugs, because there was no checking on these buffers. Tsk,
3939 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3940 buffer is too small.
3942 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3943 lsearch lookups. Now it does.
3945 46. When parsing a route_list item in a manualroute router, a fixed-length
3946 buffer was used for the list of hosts. I made this 1024 bytes long,
3947 thinking that nobody would ever have a list of hosts that long. Wrong.
3948 Somebody had a whole pile of complicated expansion conditions, and the
3949 string was silently truncated, leading to an expansion error. It turns out
3950 that it is easier to change to an unlimited length (owing to other changes
3951 that have happened since this code was originally written) than to build
3952 structure for giving a limitation error. The length of the item that
3953 expands into the list of hosts is now unlimited.
3955 47. The lsearch lookup could not handle data where the length of text line was
3956 more than 4095 characters. Such lines were truncated, leading to shortened
3957 data being returned. It should now handle lines of any length.
3959 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3960 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3963 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3964 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3965 no longer happens in comment lines. A list of replacements is now placed
3966 at the head of all of the source files, except those whose only change is
3967 to replace PERL_COMMAND in the very first #! line.
3969 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3970 messages on the queue, with a bottom-up merge sort, using code contributed
3971 by Michael Haardt. This should make operations like -bp somewhat faster on
3972 large queues. It won't affect queue runners, except when queue_run_in_order
3975 51. Installed eximstats 1.31 in the distribution.
3977 52. Added support for SRV lookups to the dnslookup router.
3979 53. If an ACL referred to $message_body or $message_body_end, the value was not
3980 reset for any messages that followed in the same SMTP session.
3982 54. The store-handling optimization for building very long strings was not
3983 differentiating between the different store pools. I don't think this
3984 actually made any difference in practice, but I've tidied it.
3986 55. While running the routers to verify a sender address, $sender_address
3987 was still set to the sender address. This is wrong, because when routing to
3988 send a bounce to the sender, it would be empty. Therefore, I have changed
3989 it so that, while verifying a sender address, $sender_address is set to <>.
3990 (There is no change to what happens when verifying a recipient address.)
3992 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3993 target A or AAAA records (if not already returned) without resetting the
3994 qualify_single or search_parents options of the DNS resolver. These are
3995 inappropriate in this case because the targets of MX and SRV records must
3996 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3997 target that, when qualified, matched something in the local domain. These
3998 two options are now turned off when doing these lookups.
4000 57. It seems that at least some releases of Reiserfs (which does not have the
4001 concept of a fixed number of inodes) returns zero and not -1 for the
4002 number of available inodes. This interacted badly with check_spool_inodes,
4003 which assumed that -1 was the "no such thing" setting. What I have done is
4004 to check that the total number of inodes is greater than zero before doing
4005 the test of how many are available.
4007 58. When a "warn" ACL statement has a log_message modifier, the message is
4008 remembered, and not repeated. This is to avoid a lot of repetition when a
4009 message has many recipients that cause the same warning to be written.
4010 Howewer, Exim was preserving the list of already written lines for an
4011 entire SMTP session, which doesn't seem right. The memory is now reset if a
4012 new message is started.
4014 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4015 debugging output unless log_rewrite was also set.
4017 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4018 of (int)(handle) when we know that handle contains (void *)(-1).
4020 61. The Exim daemon panic-logs an error return when it closes the incoming
4021 connection. However "connection reset by peer" seems to be common, and
4022 isn't really an error worthy of noting specially, so that particular error
4025 62. When Exim is trying to find all the local interfaces, it used to panic and
4026 die if the ioctl to get the interface flags failed. However, it seems that
4027 on at least one OS (Solaris 9) it is possible to have an interface that is
4028 included in the list of interfaces, but for which you get a failure error
4029 for this call. This happens when the interface is not "plumbed" into a
4030 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4031 failure of the "get flags" call assumes that the interface is down.
4033 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4034 makes life easier for people who are doing arithmetic on fields extracted
4035 from dates, where you often get leading zeros that should not be
4036 interpreted as octal.
4038 64. Added qualify_domain to the redirect router, to override the global
4041 65. If a pathologically long header line contained very many addresses (the
4042 report of this problem mentioned 10 000) and each of them was rewritten,
4043 Exim could use up a very large amount of memory. (It kept on making new
4044 copies of the header line as it rewrote, and never released the old ones.)
4045 At the expense of a bit more processing, the header rewriting function has
4046 been changed so that it no longer eats memory in this way.
4048 66. The generation of the Received: header has been moved from the time that a
4049 message starts to be received, to the time that it finishes. The timestamp
4050 in the Received: header should now be very close to that of the <= log
4051 line. There are two side-effects of this change:
4053 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4054 logged header lines no longer include the local Received: line, because
4055 it has not yet been created. The same applies to a copy of the message
4056 that is returned to a non-SMTP sender when a message is rejected.
4058 (b) When a filter file is tested using -bf, no additional Received: header
4059 is added to the test message. After some thought, I decided that this
4062 This change does not affect the value of $received_for. It is still set
4063 after address rewriting, but before local_scan() is called.
4065 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4067 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4068 gave an unhelpful panic error message, and a defer error. I have managed to
4069 change this behaviour so that it now rejects any supplied certificate,
4070 which seems right, as the list of acceptable certificates is empty.
4072 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4073 gave an unhelpful defer error. I have not managed to make this reject any
4074 supplied certificates, but the error message it gives is "no certificate
4075 supplied", which is not helpful.
4077 70. exigrep's output now also includes lines that are not associated with any
4078 message, but which match the given pattern. Implemented by a patch from
4079 Martin Sluka, which also tidied up the Perl a bit.
4081 71. Recipient callout verification, like sender verification, was using <> in
4082 the MAIL FROM command. This isn't really the right thing, since the actual
4083 sender may affect whether the remote host accepts the recipient or not. I
4084 have changed it to use the actual sender in the callout; this means that
4085 the cache record is now keyed on a recipient/sender pair, not just the
4086 recipient address. There doesn't seem to be a real danger of callout loops,
4087 since a callout by the remote host to check the sender would use <>.
4088 [SEE ABOVE: changed after hitting problems.]
4090 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4091 temporary errors. However, in the case of such a code being given after
4092 the end of a data transmission (i.e. after ".") Exim was failing to write
4093 a retry record for the message. (Yes, there was some broken host that was
4094 actually sending 8xx at this point.)
4096 73. An unknown lookup type in a host list could cause Exim to panic-die when
4097 the list was checked. (An example that provoked this was putting <; in the
4098 middle of a list instead of at the start.) If this happened during a DATA
4099 ACL check, a -D file could be left lying around. This kind of configuration
4100 error no longer causes Exim to die; instead it causes a defer errror. The
4101 incident is still logged to the main and panic logs.
4103 74. Buglet left over from Exim 3 conversion. The message "too many messages
4104 in one connection" was written to the rejectlog but not the mainlog, except
4105 when address rewriting (yes!) was being logged.
4107 75. Added write_rejectlog option.
4109 76. When a system filter was run not as root (that is, when system_filter_user
4110 was set), the values of the $n variables were not being returned to the
4111 main process; thus, they were not subsequently available in the $sn
4114 77. Added +return_path_on_delivery log selector.
4116 78. A connection timeout was being treated differently from recipients deferred
4117 when testing hosts_max_try with a message that was older than the host's
4118 retry timeout. (The host should not be counted, thus allowing all hosts to
4119 be tried at least once before bouncing.) This may have been the cause of an
4120 occasionally reported bug whereby a message would remain on the queue
4121 longer than the retry timeout, but would be bounced if a delivery was
4122 forced. I say "may" because I never totally pinned down the problem;
4123 setting up timeout/retry tests is difficult. See also the next item.
4125 79. The ultimate address timeout was not being applied to errors that involved
4126 a combination of host plus message (for example, a timeout on a MAIL
4127 command). When an address resolved to a number of possible hosts, and they
4128 were not all tried for each delivery (e.g. because of hosts_max_try), a
4129 message could remain on the queue longer than the retry timeout.
4131 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4134 81. Fixed an obscure SMTP outgoing bug which required at least the following
4135 conditions: (a) there was another message waiting for the same server;
4136 (b) the server returned 5xx to all RCPT commands in the first message so
4137 that the message was not completed; (c) the server dropped the connection
4138 or gave a negative response to the RSET that Exim sends to abort the
4139 transaction. The observed case was a dropped connection after DATA that had
4140 been sent in pipelining mode. That is, the server had advertised PIPELINING
4141 but was not implementing it correctly. The effect of the bug was incorrect
4142 behaviour, such as trying another host, and this could lead to a crash.
4148 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4149 and daemon.c were passed as pointers to ints; they should have been
4150 pointers to socklen_t variables (which are typically unsigned ints).
4152 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4155 3. Fixed a really odd bug that affected only the testing scheme; patching a
4156 certain fixed string in the binary changed the value of another string that
4157 happened to be identical to the end of the original first string.
4159 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4160 name", it returns that address as the IP address. On some operating
4161 systems (e.g. Solaris), it also passes back the IP address string as the
4162 "host name". However, on others (e.g. Linux), it passes back an empty
4163 string. Exim wasn't checking for this, and was changing the host name to an
4164 empty string, assuming it had been canonicized.
4166 5. Although rare, it is permitted to have more than one PTR record for a given
4167 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4168 all the names associated with an address, because they do in Solaris.
4169 However, it seems that they do not in Linux for data that comes from the
4170 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4171 I found this out when I moved to a new Linux workstation and tried to run
4172 the Exim test suite.
4174 To get round this problem I have changed the code so that it now does its
4175 own call to the DNS to look up PTR records when searching for a host name.
4176 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4177 addresses that are only in /etc/hosts are still found.
4179 This behaviour is, however, controlled by an option called host_lookup_
4180 order, which defaults to "bydns:byaddr". If people want to use the other
4181 order, or indeed, just use one or the other means of lookup, they can
4182 specify it in this variable.
4184 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4185 some operating systems, this comes back from gethostbyaddr() as an empty
4186 string, and this is what Exim used to test for. However, it seems that in
4187 other systems, "." is yielded. Exim now tests for this case too.
4189 7. The values of check_spool_space and check_log_space are now held internally
4190 as a number of kilobytes instead of an absolute number of bytes. If a
4191 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4192 kilobyte. This means that much larger values can be stored.
4194 8. Exim monitor: an attempt to get the action menu when not actually pointing
4195 at a message produces an empty menu entitled "No message selected". This
4196 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4197 no entries in it ("Shell widget menu has zero width and/or height"). So I
4198 have added a single, blank menu entry in this case.
4200 9. Added ${quote_local_part.
4202 10. MIME decoding is now applied to the contents of Subject: header lines when
4205 11. Now that a reference to $sender_host_address automatically causes a reverse
4206 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4207 host lookup before query-style lookups in lists that might use this
4208 variable. This has therefore been abolished, and the "net-" prefix is no
4209 longer necessary for query-style lookups.
4211 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4212 have been a typo for LFLAGS, so it has been changed.
4214 13. The install script calls Exim with "-C /dev/null" in order to find the
4215 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4216 to be output. Howeve, since Exim outputs its version number before the
4217 error, it didn't break the script. It just looked ugly. I fixed this by
4218 always allowing "-C /dev/null" if the caller is root.
4220 14. Ignore overlarge ACL variable number when reading spool file - insurance
4221 against a later release with more variables having written the file.
4223 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4224 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4225 this, as well as the form without the "IPv6" on the front.
4227 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4228 OS/Makefile-Darwin file.
4230 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4231 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4233 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4234 variable. However, some OS use ints for inodes. I've added cast to long int
4235 to get rid of the compiler warning.
4237 19. I had forgotten to lock out "/../" in configuration file names when
4238 ALT_CONFIG_PREFIX was set.
4240 20. Routers used for verification do not need to specify transports. However,
4241 if such a router generated a host list, and callout was configured, Exim
4242 crashed, because it could not find a port number from the (non-existent)
4243 transport. It now assumes port 25 in this circumstance.
4245 21. Added the -t option to exigrep.
4247 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4248 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4249 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4250 them is set, LOOKUP_LSEARCH is forced.
4252 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4254 24. Added sender and host information to the "rejected by local_scan()" log
4255 line; previously there was no indication of these.
4257 25. Added .include_if_exists.
4259 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4260 Linux. It turns out that not all file systems support this. Apparently some
4261 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4262 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4263 not supported on the file descriptor, is now ignored when Exim is trying to
4264 sync a directory. This applies only to Linux.
4266 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4268 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4269 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4272 29. Added support for maildirsize files from supplied patch (modified a bit).
4274 30. The use of :fail: followed by an empty string could lead Exim to respond to
4275 sender verification failures with (e.g.):
4277 550 Verification failed for <xxx>
4278 550 Sender verify failed
4280 where the first response line was missing the '-' that indicates it is not
4281 the final line of the response.
4283 31. The loop for finding the name of the user that called Exim had a hardwired
4284 limit of 10; it now uses the value of finduser_retries, which is used for
4285 all other user lookups.
4287 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4290 33. Exim was neglecting to zero errno before one call of strtol() when
4291 expanding a string and expecting an integer value. On some systems this
4292 resulted in spurious "integer overflow" errors. Also, it was casting the
4293 result into an int without checking.
4295 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4296 did not work. The code looks as if it has *never* worked, though it appears
4297 to have been documented since at least releast 1.62. I have made it work.
4299 35. The "timeout_DNS" error in retry rules, also documented since at least
4300 1.62, also never worked. As it isn't clear exactly what this means, and
4301 clearly it isn't a major issue, I have abolished the feature by treating it
4302 as "timeout", and writing a warning to the main and panic logs.
4304 36. The display of retry rules for -brt wasn't always showing the error code
4307 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4308 timeout_connect_A, timeout_connect_MX.
4310 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4311 to the empty sender.
4313 39. The daemon was not analysing the content of -oX till after it had closed
4314 stderr and disconnected from the controlling terminal. This meant that any
4315 syntax errors were only noted on the panic log, and the return code from
4316 the command was 0. By re-arranging the code a little, I've made the
4317 decoding happen first, so such errors now appear on stderr, and the return
4318 code is 1. However, the actual setting up of the sockets still happens in
4319 the disconnected process, so errors there are still only recorded on the
4322 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4323 connections (as happens on some IP stacks) was logged at start up time as
4324 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4325 it from "IPv6 and IPv4", which means that two separate sockets are being
4328 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4329 says whether AF_INET or AF_INET6 was passed as an argument.
4331 42. Exiwhat output was messed up when time zones were included in log
4334 43. Exiwhat now gives more information about the daemon's listening ports,
4335 and whether -tls-on-connect was used.
4337 44. The "port" option of the smtp transport is now expanded.
4339 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4340 silently ignored. Now an error message is written to the main and panic
4343 46. There's a new ACL modifier called "logwrite" which writes to a log file
4344 as soon as it is encountered.
4346 47. Added $local_user_uid and $local_user_gid at routing time.
4348 48. Exim crashed when trying to verify a sender address that was being
4351 49. Exim was recognizing only a space character after ".include". It now also
4352 recognizes a tab character.
4354 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4355 extracting the relevant information from the specification. The man page no
4356 longer contains scrambled data for the -d option, and I've added a section
4357 at the front about calling Exim under different names.
4359 51. Added "extra_headers" argument to the "mail" command in filter files.
4361 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4364 53. Installed eximstats 1.29.
4366 54. Added transport_filter_timeout as a generic transport option.
4368 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4369 Cc: header lines. This was required by RFC 822, but it not required by RFC
4372 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4373 incoming messages that did not have them. Now it does so only if the
4374 message originates locally, that is, if there is no associated remote host
4375 address. When Resent- header lines are present, this applies to the Resent-
4376 lines rather than the non-Resent- lines.
4378 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4379 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4381 58. Messages for configuration errors now include the name of the main
4382 configuration file - useful now that there may be more than one file in a
4383 list (.included file names were always shown).
4385 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4386 for those rare installations that do not start the daemon as root or run it
4387 setuid root. I've cut out the call to initgroups() if the daemon is not
4390 60. The Exim user and group can now be bound into the binary as text strings
4391 that are looked up at the start of Exim's processing.
4393 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4395 62. Added $mailstore_basename variable.
4397 63. Installed patch to sieve.c from Michael Haardt.
4399 64. When Exim failed to open the panic log after failing to open the main log,
4400 the original message it was trying to log was written to stderr and debug
4401 output, but if they were not available (the usual case in production), it
4402 was lost. Now it is written to syslog before the two lines that record the
4403 failures to open the logs.
4405 65. Users' Exim filters run in subprocesses under the user's uid. It is
4406 possible for a "deliver" command or an alias in a "personal" command to
4407 provoke an address rewrite. If logging of address rewriting is configured,
4408 this fails because the process is not running as root or exim. There may be
4409 a better way of dealing with this, but for the moment (because 4.30 needs
4410 to be released), I have disabled address rewrite logging when running a
4411 filter in a non-root, non-exim process.
4417 1. The buildconfig auxiliary program wasn't quoting the value set for
4418 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4419 not defined. This bug was masked in 4.22 by the effect that was fixed in
4422 2. Some messages that were rejected after a message id was allocated were
4423 shown as "incomplete" by exigrep. It no longer does this for messages that
4424 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4426 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4427 have allow_domain_literals set, the ID did not get logged in the <= line.
4428 Domain literals are now always recognized in Message-ID: header lines.
4430 4. The first argument for a ${extract expansion item is the key name or field
4431 number. Leading and trailing spaces in this item were not being ignored,
4432 causing some misleading effects.
4434 5. When deliver_drop_privilege was set, single queue runner processes started
4435 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4436 same command in the process it spins off) were not dropping privilege.
4438 6. When the daemon running as "exim" started a queue runner, it always
4439 re-executed Exim in the spun-off process. This is a waste of effort when
4440 deliver_drop_privilege is set. The new process now just calls the
4441 queue-runner function directly.
4447 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4450 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4451 ignored. Though the use of -oP was forcing the writing of a pid file, it
4452 was always written to the default place.
4454 3. If the message "no IP address found for host xxxx" is generated during
4455 incoming verification, it is now followed by identification of the incoming
4456 connection (so you can more easily find what provoked it).
4458 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4460 5. Added some features to "harden" Exim a bit more against certain attacks:
4462 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4463 be put in Local/Makefile. This is like the never_users runtime option,
4464 but it cannot be overridden. The default setting is "root".
4466 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4467 prefix string with which any file named in a -C command line option
4470 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4471 is retained for -C and -D only if the caller of Exim is root. Without
4472 it, the exim user may also use -C and -D and retain privilege.
4474 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4475 command line option is disabled.
4477 6. Macro names set by the -D option must start with an upper case letter, just
4478 like macro names defined in the configuration file.
4480 7. Added "dereference=" facility to LDAP.
4482 8. Two instances of the typo "uknown" in the source files are fixed.
4484 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4485 the Configure-Makefile script screwed up while processing it.
4487 10. Incorporated PCRE 4.4.
4489 11. The SMTP synchronization check was not operating right at the start of an
4490 SMTP session. For example, it could not catch a HELO sent before the client
4491 waited for the greeting. There is now a check for outstanding input at the
4492 point when the greeting is written. Because of the duplex, asynchronous
4493 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4494 way, but not yet received, when the check is performed.
4496 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4497 on TCP/IP sockets, because this apparently causes some broken clients to
4500 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4501 unchanged) from the Cygwin maintainer.
4503 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4504 was supported instead of testing for mbx. Effectively a typo.
4506 15. The spa authenticator server code was not checking that the input it
4507 received was valid base64.
4509 16. The debug output line for the "set" modifier in ACLs was not showing the
4510 name of the variable that was being set.
4512 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4514 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4515 lookup on its own. Something else was needed to trigger the lookup. For
4516 example, a match in host_lookup or the need for a host name in a host list.
4517 Now, if $sender_host_name is referenced and the host name has not yet been
4518 looked up, a lookup is performed. If the lookup fails, the variable remains
4519 empty, and $host_lookup_failed is set to "1".
4521 19. Added "eqi" as a case-independent comparison operator.
4523 20. The saslauthd authentication condition could segfault if neither service
4524 nor realm was specified.
4526 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4527 error message that was logged was misleading, and incoming SMTP
4528 connections were dropped. The message is now more accurate, and temporary
4529 errors are given to SMTP connections.
4531 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4532 (except RSET) with a 421 error, until QUIT is received. However, it was
4533 failing to send a response to QUIT.
4535 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4536 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4537 ACL is now run after the helo_try_verify_hosts code.
4539 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4540 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4541 case-independent, so other case variants are also recognized). Apparently
4542 some systems use these upper case variants.
4544 25. If more than two messages were waiting for the same host, and a transport
4545 filter was specified for the transport, Exim sent two messages over the
4546 same TCP/IP connection, and then failed with "socket operation on non-
4547 socket" when it tried to send the third.
4549 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4551 27. The extern definition of crypt16() in expand.c was not being excluded when
4552 the OS had its own crypt16() function.
4554 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4555 as a preferred synonym for return_size_limit, both as an option and as an
4558 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4560 30. Changed the default configuration ACL to relax the local part checking rule
4561 for addresses that are not in any local domains. For these addresses,
4562 slashes and pipe symbols are allowed within local parts, but the sequence
4563 /../ is explicitly forbidden.
4565 31. SPA server authentication was not clearing the challenge buffer before
4568 32. log_message in a "warn" ACL statement was writing to the reject log as
4569 well as to the main log, which contradicts the documentation and doesn't
4570 seem right (because no rejection is happening). So I have stopped it.
4572 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4573 However, I am unable to do any testing of this.
4575 34. Fixed an infelicity in the appendfile transport. When checking directories
4576 for a mailbox, to see if any needed to be created, it was accidentally
4577 using path names with one or more superfluous leading slashes; tracing
4578 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4580 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4581 contents are added to the log line that is written for every discarded
4582 recipient. (Previously a log_message setting was ignored.)
4584 36. The ${quote: operator now quotes the string if it is empty.
4586 37. The install script runs exim in order to find its version number. If for
4587 some reason other than non-existence or emptiness, which it checks, it
4588 could not run './exim', it was installing it with an empty version number,
4589 i.e. as "exim-". This error state is now caught, and the installation is
4592 38. An argument was missing from the function that creates an error message
4593 when Exim fails to connect to the socket for saslauthd authentication.
4594 This could cause Exim to crash, or give a corrupted message.
4596 39. Added isip, isip4, and isip6 to ${if conditions.
4598 40. The ACL variables $acl_xx are now saved with the message, and can be
4599 accessed later in routers, transports, and filters.
4601 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4602 strings in the file are not string-expanded.
4604 42. If a MAIL command specified a SIZE value that was too large to fit into an
4605 int variable, the check against message_size_limit failed. Such values are
4606 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4607 day this will have to be increased, but I don't think I want to be around
4608 when emails are that large.
4615 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4616 iconv() is not standard in FreeBSD.
4618 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4619 IPv6 enabled. The observed symptom was a segmentation fault on return from
4620 the function os_common_find_running_interfaces() in src/os.c.
4622 3. In the check_special_case() function in daemon.c I had used "errno" as an
4623 argument name, which causes warnings on some systems. This was basically a
4624 typo, since it was named "eno" in the comments!
4626 4. The code that waits for the clock to tick (at a resolution of some fraction
4627 of a second) so as to ensure message-id uniqueness was always waiting for
4628 at least one whole tick, when it could have waited for less. [This is
4629 almost certainly not relevant at current processor speeds, where it is
4630 unlikely to ever wait at all. But we try to future-proof.]
4632 5. The function that sleeps for a time interval that includes fractions of a
4633 second contained a race. It did not block SIGALRM between setting the
4634 timer, and suspending (a couple of lines later). If the interval was short
4635 and the sigsuspend() was delayed until after it had expired, the suspension
4636 never ended. On busy systems this could lead to processes getting stuck for
4639 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4640 process, before it forks any delivery processes. The open lookup caching
4641 mechanism meant that the open file or database connection was passed into
4642 the delivery process. The problem was that delivery processes always tidy
4643 up cached lookup data. This could cause a problem for the next delivery
4644 process started by the queue runner, because the external queue runner
4645 process does not know about the closure. So the next delivery process
4646 still has data in the lookup cache. In the case of a file lookup, there was
4647 no problem because closing a file descriptor in a subprocess doesn't affect
4648 the parent. However, if the lookup was caching a connection to a database,
4649 the connection was closed, and the second delivery process was likely to
4650 see errors such as "PGSQL: query failed: server closed the connection
4651 unexpectedly". The problem has been fixed by closing all cached lookups
4652 in a queue runner before running a delivery process.
4654 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4655 seem to have the "const" qualifier which it has on other OS. I've
4658 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4659 *of the same type* (client or server) with the same public name that an
4660 error should be diagnosed.
4662 9. When Exim looked up a host name for an IP address, but failed to find the
4663 original IP address when looking up the host name (a safety check), it
4664 output the message "<ip address> does not match any IP for NULL", which was
4665 confusing, to say the least. The bug was that the host name should have
4666 appeared instead of "NULL".
4668 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4669 user that is built into the binary, and the -C or -D options is used, root
4670 privilege is dropped before the configuration file is read. In addition,
4671 logging is switched to stderr instead of the normal log files. If the
4672 configuration then re-defines the Exim user, the unprivileged environment
4673 is probably not what is expected, so Exim logs a panic warning message (but
4676 However, if deliver_drop_privilege is set, the unprivileged state may well
4677 be exactly what is intended, so the warning has been cut out in that case,
4678 and Exim is allowed to try to write to its normal log files.
4684 1. smtp_return_error_details was not giving details for temporary sender
4685 or receiver verification errors.
4687 2. Diagnose a configuration error if two authenticators have the same public
4690 3. Exim used not to create the message log file for a message until the first
4691 delivery attempt. This could be confusing when incoming messages were held
4692 for policy or load reasons. The message log file is now created at the time
4693 the message is received, and an initial "Received" line is written to it.
4695 4. The automatically generated man page for command line options had a minor
4696 bug that caused no ill effects; however, a more serious problem was that
4697 the procedure for building the man page automatically didn't always
4698 operate. Consequently, release 4.20 contains an out-of-date version. This
4699 shouldn't happen again.
4701 5. When building Exim with embedded Perl support, the script that builds the
4702 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4703 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4705 6. The freeze_tell option was not being used for messages that were frozen on
4706 arrival, either by an ACL or by local_scan().
4708 7. Added the smtp_incomplete_transaction log selector.
4710 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4711 was accepting AUTH without a new EHLO.
4713 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4714 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4717 10. Logging of TCP/IP connections (when configured) now happens in the main
4718 daemon process instead of the child process, so that the TCP/IP connection
4719 count is more accurate (but it can never be perfect).
4721 11. The use of "drop" in a nested ACL was not being handled correctly in the
4722 outer ACL. Now, if condition failure induced by the nested "drop" causes
4723 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4724 or "require"), the connection is dropped.
4726 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4727 that yield "discard" can now be used with an "accept" or a "discard" verb,
4728 but an error is generated for any others (because I can't see a useful way
4729 to define what should happen).
4731 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4732 are now processed in the same way as lines in the Exim configuration file.
4733 In particular, continuation lines are supported.
4735 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4737 15. Added -ti meaning -t -i.
4739 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4740 domains, and warn by logging if others are found.
4742 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4743 returned from GIFCONF when Exim is trying to find the list of interfaces on
4744 a host. The code in os.c has been modified to copy each ifreq to an aligned
4745 structure in all cases.
4747 Also, in some cases, the returned ifreq's were being copied to a 'struct
4748 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4749 means the last couple of bytes of an IPv6 address could be chopped if the
4750 ifreq contained only a normal sockaddr (14 bytes storage).
4752 18. Named domain lists were not supported in the hosts_treat_as_local option.
4753 An entry such as +xxxx was not recognized, and was treated as a literal
4756 19. Ensure that header lines added by a DATA ACL are included in the reject log
4757 if the ACL subsequently rejects the message.
4759 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4760 MD5 (which is deprecated).
4762 21. When testing a filter file using -bf, Exim was writing a message when it
4763 took the sender from a "From " line in the message, but it was not doing so
4764 when it took $return_path from a Return-Path: header line. It now does.
4766 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4767 with a valid header line field name (a series of printing characters
4768 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4770 23. Changed "disc" in the source to "disk" to conform to the documentation and
4771 the book and for uniformity.
4773 24. Ignore Sendmail's -Ooption=value command line item.
4775 25. When execve() failed while trying to run a command in a pipe transport,
4776 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4777 could be confused with a return value of 69 from the command itself. This
4778 has been changed to 127, the value the shell returns if it is asked to run
4779 a non-existent command. The wording for the related log line suggests a
4780 non-existent command as the problem.
4782 26. If received_header_text expands to an empty string, do not add a Received:
4783 header line to the message. (Well, it adds a token one on the spool, but
4784 marks it "old" so that it doesn't get used or transmitted.)
4786 27. Installed eximstats 1.28 (addition of -nt option).
4788 28. There was no check for failure on the call to getsockname() in the daemon
4789 code. This can fail if there is a shortage of resources on the system, with
4790 ENOMEM, for example. A temporary error is now given on failure.
4792 29. Contrary to the C standard, it seems that in some environments, the
4793 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4794 program. Exim now does this explicitly; it affects the formatting of
4795 timestamps using strftime().
4797 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4798 complaints. I've now initialized all the variables, to avoid this.
4800 32. Header lines added by a system filter were not being "seen" during
4801 transport-time rewrites.
4803 33. The info_callback() function passed to OpenSSL is set up with type void
4804 (*)(SSL *, int, int), as described somewhere. However, when calling the
4805 function (actually a macro) that sets it up, the type void(*)() is
4806 expected. I've put in a cast to prevent warnings from picky compilers.
4808 34. If a DNS black list lookup found a CNAME record, but there were no A
4809 records associated with the domain it pointed at, Exim crashed.
4811 35. If a DNS black list lookup returned more than one A record, Exim ignored
4812 all but the first. It now scans all returned addresses if a particular IP
4813 value is being sought. In this situation, the contents of the
4814 $dnslist_value variable are a list of all the addresses, separated by a
4817 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4818 to accept a host name and all its aliases if the forward lookup for any of
4819 them yielded the IP address of the incoming connection. Now it accepts only
4820 those names whose forward lookup yields the correct IP address. Any other
4821 names are discarded. This closes a loophole whereby a rogue DNS
4822 administrator could create reverse DNS records to break through a
4823 wildcarded host restriction in an ACL.
4825 37. If a user filter or a system filter that ran in a subprocess used any of
4826 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4827 the wrong values were passed to the pipe command ($thisaddress had the
4828 value of $0, $0 had the value of $1, etc). This bug was introduced by
4829 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4831 38. Improved the line breaking for long SMTP error messages from ACLs.
4832 Previously, if there was no break point between 40 and 75 characters, Exim
4833 left the rest of the message alone. Two changes have been made: (a) I've
4834 reduced the minimum length to 35 characters; (b) if it can't find a break
4835 point between 35 and 75 characters, it looks ahead and uses the first one
4836 that it finds. This may give the occasional overlong line, but at least the
4837 remaining text gets split now.
4839 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4840 file descriptors might be low, and that setting 1000 would always raise it.
4841 It turns out that in some environments, the limit is already over 1000 and
4842 that lowering it causes trouble. So now Exim takes care not to decrease it.
4844 40. When delivering a message, the value of $return_path is set to $sender_
4845 address at the start of routing (routers may change the value). By an
4846 oversight, this default was not being set up when an address was tested by
4847 -bt or -bv, which affected the outcome if any router or filter referred to
4850 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4851 log only when "message" or "log_message" are set. However, if one of the
4852 conditions was an address verification, or a call to a nested ACL, the
4853 messages generated by the underlying test were being passed through. This
4854 no longer happens. The underlying message is available in $acl_verify_
4855 message for both "message" and "log_message" expansions, so it can be
4856 passed through if needed.
4858 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4859 new expansion $bh_ to give the encoded byte string without charset
4860 translation. Translation happens only if iconv() is available; HAVE_ICONV
4861 indicates this at build time. HEADERS_CHARSET gives the charset to
4862 translate to; headers_charset can change it in the configuration, and
4863 "headers charset" can change it in an individual filter file.
4865 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4866 that creates RFC 2047 encoded "words" labels them as that charset instead
4867 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4868 expansion operator; (ii) when Exim creates a From: line for a local
4869 message; (iii) when a header line is rewritten to include a "phrase" part.
4871 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4872 buggy, causing it to skip the first lines of messages whose message ID
4873 ended in 'D'. This would not have bitten before Exim release 4.14, because
4874 message IDs were unlikely to end in 'D' before then. The effect was to have
4875 incorrect size information for certain domains.
4877 45. #include "config.h" was missing at the start of the crypt16.c module. This
4878 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4881 46. If there was a timeout during a "random" callout check, Exim treated it as
4882 a failure of the random address, and carried on sending RSET and the real
4883 address. If the delay was just some slowness somewhere, the response to the
4884 original RCPT would be taken as a response to RSET and so on, causing
4885 mayhem of various kinds.
4887 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4888 when I implemented it. It didn't allow for the fact that some option values
4889 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4890 the right test for positive values.
4892 48. Domain names in DNS records are case-independent. Exim always looks them up
4893 in lower case. Some resolvers return domain names in exactly the case they
4894 appear in the zone file, that is, they may contain uppercase letters. Not
4895 all resolvers do this - some return always lower case. Exim was treating a
4896 change of case by a resolver as a change of domain, similar to a widening
4897 of a domain abbreviation. This triggered its re-routing code and so it was
4898 trying to route what was effectively the same domain again. This normally
4899 caused routing to fail (because the router wouldn't handle the domain
4900 twice). Now Exim checks for this case specially, and just changes the
4901 casing of the domain that it ultimately uses when it transmits the message
4904 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4907 50. If a filter generated a file delivery with a non-absolute name (possible if
4908 no home directory exists for the router), the forbid_file option was not
4911 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4912 the existing equality matching.
4914 52. Exim was using ints instead of ino_t variables in some places where it was
4915 dealing with inode numbers.
4917 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4918 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4919 called TMPDIR, and if it finds it is different, it changes its value.
4921 54. The smtp_printf() function is now made available to local_scan() so
4922 additional output lines can be written before returning. There is also an
4923 smtp_fflush() function to enable the detection of a dropped connection.
4924 The variables smtp_input and smtp_batched_input are exported to
4927 55. Changed the default runtime configuration: the message "Unknown user"
4928 has been removed from the ACL, and instead placed on the localuser router,
4929 using the cannot_route_message feature. This means that any verification
4930 failures that generate their own messages won't get overridden. Similarly,
4931 the "Unrouteable address" message that was in the ACL for unverifiable
4932 relay addresses has also been removed.
4934 56. Added hosts_avoid_esmtp to the smtp transport.
4936 57. The exicyclog script was not checking for the esoteric option
4937 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4938 will work only if exicyclog is run under the appropriate euid.
4940 58. Following a discussion on the list, the rules by which Exim recognises line
4941 endings on incoming messages have been changed. The -dropcr and drop_cr
4942 options are now no-ops, retained only for backwards compatibility. The
4943 following line terminators are recognized: LF CRLF CR. However, special
4944 processing applies to CR:
4946 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4947 nor a local message in the state where . is a terminator.
4949 (ii) If a bare CR is encountered in a header line, an extra space is added
4950 after the line terminator so as not to end the header. The reasoning
4951 behind this is that bare CRs in header lines are most likely either
4952 to be mistakes, or people trying to play silly games.
4954 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4955 was being incorrectly given as 18 bytes larger than it should have been.
4956 This is a VOB (very old bug).
4958 60. This may never have affected anything current, but just in case it has:
4959 When the local host is found other than at the start of a list of hosts,
4960 the local host, those with the same MX, and any that follow, are discarded.
4961 When the list in question was part of a longer list of hosts, the following
4962 hosts (not currently being processed) were also being discarded. This no
4963 longer happens. I'm not sure if this situation could ever has previously
4966 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4969 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4970 header. This is something that is recommended in a new Internet Draft, and
4971 is something that is documented as being done by Sendmail. There are two
4972 possible values. For messages generated by the autoreply transport, Exim
4975 Auto-Submitted: auto-replied
4977 whereas for all other generated messages (e.g. bounces) it adds
4979 Auto-Submitted: auto-generated
4981 63. The "personal" condition in filters now includes a test for the
4982 Auto-Submitted: header. If it contains the string "auto-" the message it
4983 not considered personal.
4985 64. Added rcpt_include_affixes as a generic transport option.
4987 65. Added queue_only_override (default true).
4989 66. Added the syslog_duplication option.
4991 67. If what should have been the first header line of a message consisted of
4992 a space followed by a colon, Exim was mis-interpreting it as a header line.
4993 It isn't of course - it is syntactically invalid and should therefore be
4994 treated as the start of the message body. The misbehaviour could have
4995 caused a number of strange effects, including loss of data in subsequent
4996 header lines, and spool format errors.
4998 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4999 client host had authenticated. This control can now be exercised by an ACL
5000 for more flexibility.
5002 69. By default, callouts do not happen when testing with -bh. There is now a
5003 variant, -bhc, which does actually run the callout code, including
5004 consulting and updating the callout cache.
5006 70. Added support for saslauthd authentication, courtesy of Alexander
5009 71. If statvfs() failed on the spool or log directories while checking their
5010 size for availability, Exim confusingly gave the error "space shortage".
5011 Furthermore, in debugging mode it crashed with a floating point exception.
5012 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5013 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5014 serious problem, Exim now writes to the main and panic logs when this
5015 happens, with details of the failure. It then refuses to accept the
5016 incoming message, giving the message "spool directory problem" or "log
5017 directory problem" with a 421 code for SMTP messages.
5019 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5020 0, 1, and 2 exist, because some OS complain for execs without them (see
5021 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5022 descriptors. However, the code omitted to check that the open succeeded,
5023 causing mysterious errors if for some reason the permissions on /dev/null
5024 got screwed. Now Exim writes a message to the main and panic logs, and
5025 bombs out if it can't open /dev/null.
5027 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5028 interact so that it is all more flexible. It is supposed to remain
5029 backwards compatible. Also added extra_local_interfaces.
5031 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5032 to bomb out with an assertion failure - to the client this appears as a
5033 connection drop. This problem occurs in the part of the code that was taken
5034 from the Samba project. Fortunately, the assertion is in a very simple
5035 function, so I have fixed this by reproducing the function inline in the
5036 one place where it is called, and arranging for authentication to fail
5037 instead of killing the process with assert().
5039 75. The SPA client code was not working when the server requested OEM rather
5040 than Unicode encoding.
5042 76. Added code to make require_files with a specific uid setting more usable in
5043 the case where statting the file as root fails - usually a non-root-mounted
5044 NFS file system. When this happens and the failure is EACCES, Exim now
5045 forks a subprocess and does the per-uid checking as the relevant uid.
5047 77. Added process_log_path.
5049 78. If log_file_path was not explicitly set, a setting of check_log_space or
5050 check_log_inodes was ignored.
5052 79. If a space check for the spool or log partitions fails, the incident is now
5053 logged. Of course, in the latter case the data may get lost...
5055 80. Added the %p formatting code to string_format() so that it can be used to
5056 print addresses in debug_print(). Adjusted all the address printing in the
5057 debugging in store.c to use %p rather than %d.
5059 81. There was a concern that a line of code in smtp_in.c could overflow a
5060 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5061 initially expressed, the concern was not well-founded, because trailing
5062 spaces are removed early. However, if the trailing spaces were followed by
5063 a NULL, they did not get removed, so the overflow was possible. Two fixes
5066 (a) I re-wrote the offending code in a cleaner fashion.
5067 (b) If an incoming SMTP command contains a NULL character, it is rejected
5070 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5071 runs initgroups(), so that if the Exim user is in any additional groups,
5072 they will be used during message reception.
5078 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob