1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.615 2010/06/03 08:19:13 pdp Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
8 PP/01 Date: & Message-Id: revert to normally being appended to a message,
9 only prepend for the Resent-* case. Fixes regression introduced in
10 Exim 4.70 by NM/22 for Bugzilla 607.
12 PP/02 Include check_rfc2047_length in configure.default because we're seeing
13 increasing numbers of administrators be bitten by this.
18 JJ/01 installed exipick 20100104.1, adding $max_received_linelength, $data_path, and $header_path variables; fixed documentation bugs and typos
20 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow exipick to access non-standard spools, including the "frozen" queue (Finput)
22 NM/01 Bugzilla 965: Support mysql stored procedures.
23 Patch from Alain Williams
25 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
27 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
28 Patch from Andreas Metzler
30 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
31 Patch from Kirill Miazine
33 NM/05 Bugzilla 671: Added umask to procmail example.
35 JJ/03 installed exipick 20100323.0, fixing doc bug
37 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
38 directory. Notification and patch from Dan Rosenberg.
40 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
42 TK/02 Improve log output when DKIM signing operation fails.
44 MH/01 Treat the transport option dkim_domain as a colon separated
45 list, not as a single string, and sign the message with each element,
46 omitting multiple occurences of the same signer.
48 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
49 Bugzilla 985, 986. Patch by Simon Arlott
51 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
54 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
55 MBX locking. Notification from Dan Rosenberg.
61 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body
63 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
65 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
67 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
69 NM/04 Bugzilla 915: Flag broken perl installation during build.
75 TK/01 Added patch by Johannes Berg that expands the main option
76 "spamd_address" if it starts with a dollar sign.
78 TK/02 Write list of recipients to X-Envelope-Sender header when building
79 the mbox-format spool file for content scanning (suggested by Jakob
82 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
83 (http://www.dcc-servers.net/) support via dccifd. Activated by
84 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
86 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
87 by Mark Daniel Reidel <mr@df.eu>.
89 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
90 When building exim an external PCRE library is now needed -
91 PCRE is a system library on the majority of modern systems.
92 See entry on PCRE_LIBS in EDITME file.
94 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
95 conversation. Added nologin parameter to request.
96 Patch contributed by Kirill Miazine
98 TF/01 Do not log submission mode rewrites if they do not change the address.
100 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
102 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
103 log files in place. Contributed by Roberto Lima
105 NM/04 Bugzilla 667: close socket used by dovecot authenticator
107 TF/03 Bugzilla 615: When checking the local_parts router precondition
108 after a local_part_suffix or local_part_prefix option, Exim now
109 does not use the address's named list lookup cache, since this
110 contains cached lookups for the whole local part.
112 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
113 Robert Millan. Documentation is in experimental-spec.txt
115 TF/04 Bugzilla 668: Fix parallel build (make -j).
117 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000
119 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
120 Patch provided by Jan Srzednicki
122 TF/05 Leading white space used to be stripped from $spam_report which
123 wrecked the formatting. Now it is preserved.
125 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
126 that they are available at delivery time.
128 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
130 TF/08 TLS error reporting now respects the incoming_interface and
131 incoming_port log selectors.
133 TF/09 Produce a more useful error message if an SMTP transport's hosts
134 setting expands to an empty string.
136 NM/06 Bugzilla 744: EXPN did not work under TLS.
137 Patch provided by Phil Pennock
139 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
140 Patch provided by Richard Godbee
142 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
143 acl_smtp_notquit, added index entry.
145 NM/09 Bugzilla 787: Potential buffer overflow in string_format
146 Patch provided by Eugene Bujak
148 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
149 Patch provided by Maxim Dounin
151 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
152 Patch provided by Phil Pennock
154 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
156 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
157 Patch provided by Brad "anomie" Jorsch
159 NM/14 Bugzilla 622: Added timeout setting to transport filter.
160 Patch provided by Dean Brooks
162 TK/05 Add native DKIM support (does not depend on external libraries).
164 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
165 Patch provided by Graeme Fowler
167 NM/16 Bugzilla 851: Documentation example syntax fix.
169 NM/17 Changed NOTICE file to remove references to embedded PCRE.
171 NM/18 Bugzilla 894: Fix issue with very long lines including comments in lsearch
173 NM/19 Bugzilla 745: TLS version reporting
174 Patch provided by Phil Pennock
176 NM/20 Bugzilla 167: bool: condition support
177 Patch provided by Phil Pennock
179 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken clients
180 Patch provided by Phil Pennock
182 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date
183 Patch provided by Brad "anomie" Jorsch
185 NM/23 Bugzilla 687: Fix misparses in eximstats
186 Patch provided by Heiko Schlittermann
188 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid
189 Patch provided by Heiko Schlittermann
191 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file
192 plus update to original patch.
194 NM/26 Bugzilla 799: Documentation correction for ratelimit
196 NM/27 Bugzilla 802: Improvements to local interface IP addr detection
197 Patch provided by David Brownlee
199 NM/28 Bugzilla 807: Improvements to LMTP delivery logging
201 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes
203 NM/30 Bugzilla 888: TLS documentation bugfixes
205 NM/31 Bugzilla 896: Dovecot buffer overrun fix
207 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
208 Unlike the original bugzilla I have changed all shell scripts in src tree
210 NM/33 Bugzilla 898: Transport filter timeout fix
211 Patch by Todd Rinaldo
213 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches
214 Patch by Serge Demonchaux
216 NM/35 Bugzilla 39: Base64 decode bug fixes
217 Patch by Jakob Hirsch
219 NM/36 Bugzilla 909: Correct connect() call in dcc code
221 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling
223 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed
225 NM/39 Bugzilla 911: Fixed MakeLinks build script
231 TK/01 Add preliminary DKIM support. Currently requires a forked version of
232 ALT-N's libdkim that I have put here:
233 http://duncanthrax.net/exim-experimental/
235 Note to Michael Haardt: I had to rename some vars in sieve.c. They
236 were called 'true' and it seems that C99 defines that as a reserved
237 keyword to be used with 'bool' variable types. That means you could
238 not include C99-style headers which use bools without triggering
239 build errors in sieve.c.
241 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
242 as mailq or other aliases. Changed the --help handling significantly
243 to do whats expected. exim_usage() emits usage/help information.
245 SC/01 Added the -bylocaldomain option to eximstats.
247 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
249 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
251 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
257 PH/01 Another patch from the Sieve maintainer.
259 PH/02 When an IPv6 address is converted to a string for single-key lookup
260 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
261 dots are used instead of colons so that keys in lsearch files need not
262 contain colons. This was done some time before quoting was made available
263 in lsearch files. However, iplsearch files do require colons in IPv6 keys
264 (notated using the quote facility) so as to distinguish them from IPv4
265 keys. This meant that lookups for IP addresses in host lists did not work
266 for iplsearch lookups.
268 This has been fixed by arranging for IPv6 addresses to be expressed with
269 colons if the lookup type is iplsearch. This is not incompatible, because
270 previously such lookups could never work.
272 The situation is now rather anomolous, since one *can* have colons in
273 ordinary lsearch keys. However, making the change in all cases is
274 incompatible and would probably break a number of configurations.
276 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
279 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
280 conversion specification without a maximum field width, thereby enabling
281 a rogue spamd server to cause a buffer overflow. While nobody in their
282 right mind would setup Exim to query an untrusted spamd server, an
283 attacker that gains access to a server running spamd could potentially
284 exploit this vulnerability to run arbitrary code as the Exim user.
286 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
287 $primary_hostname instead of what libspf2 thinks the hosts name is.
289 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
290 a directory entry by the name of the lookup key. Previously, if a
291 symlink pointed to a non-existing file or a file in a directory that
292 Exim lacked permissions to read, a lookup for a key matching that
293 symlink would fail. Now it is enough that a matching directory entry
294 exists, symlink or not. (Bugzilla 503.)
296 PH/03 The body_linecount and body_zerocount variables are now exported in the
299 PH/04 Added the $dnslist_matched variable.
301 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
302 This means they are set thereafter only if the connection becomes
305 PH/06 Added the client_condition to authenticators so that some can be skipped
306 by clients under certain conditions.
308 PH/07 The error message for a badly-placed control=no_multiline_responses left
309 "_responses" off the end of the name.
311 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
313 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
314 (without spaces) instead of just copying the configuration text.
316 PH/10 Added the /noupdate option to the ratelimit ACL condition.
318 PH/11 Added $max_received_linelength.
320 PH/12 Added +ignore_defer and +include_defer to host lists.
322 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
323 way in which PCRE > 7.0 is built.
325 PH/14 Implemented queue_only_load_latch.
327 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
328 MAIL command. The effect was to mangle the value on 64-bit systems.
330 PH/16 Another patch from the Sieve maintainer.
332 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
334 PH/18 If a system quota error occurred while trying to create the file for
335 a maildir delivery, the message "Mailbox is full" was not appended to the
336 bounce if the delivery eventually timed out. Change 4.67/27 below applied
337 only to a quota excession during the actual writing of the file.
339 PH/19 It seems that peer DN values may contain newlines (and other non-printing
340 characters?) which causes problems in log lines. The DN values are now
341 passed through string_printing() before being added to log lines.
343 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
344 and InterBase are left for another time.)
346 PH/21 Added message_body_newlines option.
348 PH/22 Guard against possible overflow in moan_check_errorcopy().
350 PH/23 POSIX allows open() to be a macro; guard against that.
352 PH/24 If the recipient of an error message contained an @ in the local part
353 (suitably quoted, of course), incorrect values were put in $domain and
354 $local_part during the evaluation of errors_copy.
360 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
361 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
364 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
365 issue a MAIL command.
367 PH/02 In an ACL statement such as
369 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
371 if a client was not listed at all, or was listed with a value other than
372 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
373 the condition was not true (as it should be), so access was not denied.
374 The bug was that the ! inversion was incorrectly passed on to the second
375 item. This has been fixed.
377 PH/03 Added additional dnslists conditions == and =& which are different from
378 = and & when the dns lookup returns more than one IP address.
380 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
381 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
383 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
384 FSYNC, which compiles an option called disable_fsync that allows for
385 bypassing fsync(). The documentation is heavily laced with warnings.
387 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
389 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
390 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
391 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
392 including adding "make clean"; (3) Added -fPIC when compiling the test
393 dynamically loaded module, to get rid of a warning.
395 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
396 message fails, move_frozen_messages = true and ignore_bounce_errors_after
397 = 0s. The bug is otherwise harmless.
399 PH/07 There was a bug in the dovecot authenticator such that the value of
400 $auth1 could be overwritten, and so not correctly preserved, after a
401 successful authentication. This usually meant that the value preserved by
402 the server_setid option was incorrect.
404 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
406 PH/09 Installed PCRE release 7.0.
408 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
409 run for batched SMTP input. It is now run at the start of every message
410 in the batch. While fixing this I discovered that the process information
411 (output by running exiwhat) was not always getting set for -bs and -bS
412 input. This is fixed, and it now also says "batched" for BSMTP.
414 PH/11 Added control=no_pipelining.
416 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
417 patch, slightly modified), and move the expansion of helo_data till after
418 the connection is made in the smtp transport (so it can use these
421 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
423 PH/14 Added log_selector = +pid.
425 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
427 PH/16 Add ${if forany and ${if forall.
429 PH/17 Added dsn_from option to vary the From: line in DSNs.
431 PH/18 Flush SMTP output before performing a callout, unless control =
432 no_callout_flush is set.
434 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
435 was true (the default) a successful delivery failed to delete the retry
436 item, thus causing premature timeout of the address. The bug is now
439 PH/20 Added hosts_avoid_pipelining to the smtp transport.
441 PH/21 Long custom messages for fakedefer and fakereject are now split up
442 into multiline reponses in the same way that messages for "deny" and
443 other ACL rejections are.
445 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
446 with slight modification.
448 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
449 draft, changing the syntax and factoring some duplicate code".
451 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
452 for deliveries of the second and subsequent messages over the same SMTP
455 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
456 ${reduce, with only minor "tidies".
458 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
460 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
461 expansion side effects.
463 PH/27 When a message times out after an over-quota error from an Exim-imposed
464 quota, the bounce message says "mailbox is full". This message was not
465 being given when it was a system quota that was exceeded. It now should
468 MH/03 Made $recipients available in local_scan(). local_scan() already has
469 better access to the recipient list through recipients_list[], but
470 $recipients can be useful in postmaster-provided expansion strings.
472 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
473 in the case of a MAIL command with additional options following the
474 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
475 were accidentally chopped off.
477 PH/29 SMTP synchronization checks are implemented when a command is read -
478 there is a check that no more input is waiting when there shouldn't be
479 any. However, for some commands, a delay in an ACL can mean that it is
480 some time before the response is written. In this time, more input might
481 arrive, invalidly. So now there are extra checks after an ACL has run for
482 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
483 pipelining has not been advertised.
485 PH/30 MH's patch to allow iscntrl() characters to be list separators.
487 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
488 returned in the SMTP response when smtp_return_error_details was false.
491 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
492 instead of the C I/O that was originally supplied, because problems were
495 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
496 Exim which did not show up earlier: it was assuming that a call to
497 SSL_CTX_set_info_callback() might give an error value. In fact, there is
498 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
499 was a macro that became an assignment, so it seemed to work. This has
500 changed to a proper function call with a void return, hence the compile
501 error. Exim's code has been fixed.
503 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
506 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
508 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
510 PH/37 If a message is not accepted after it has had an id assigned (e.g.
511 because it turns out to be too big or there is a timeout) there is no
512 "Completed" line in the log. When some messages of this type were
513 selected by exigrep, they were listed as "not completed". Others were
514 picked up by some special patterns. I have improved the selection
515 criteria to be more general.
517 PH/38 The host_find_failed option in the manualroute router can now be set
518 to "ignore", to completely ignore a host whose IP address cannot be
519 found. If all hosts are ignored, the behaviour is controlled by the new
520 host_all_ignored option.
522 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
523 homing or because of multiple MX records with /mx) generated more than
524 one IP address, and the following item turned out to be the local host,
525 all the secondary addresses of the first item were incorrectly removed
526 from the list, along with the local host and any following hosts (which
527 is what is supposed to happen).
529 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
530 whoever called Exim into the -H file. In the case of the daemon it was
531 behaving confusingly. When first started, it used values for whoever
532 started the daemon, but after a SIGHUP it used the Exim user (because it
533 calls itself on a restart). I have changed the code so that it now always
536 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
537 message are rejected with the same error (e.g. no authentication or bad
538 sender address), and a DATA command is nevertheless sent (as can happen
539 with PIPELINING or a stupid MUA), the error message that was given to the
540 RCPT commands is included in the rejection of the DATA command. This is
541 intended to be helpful for MUAs that show only the final error to their
544 PH/42 Another patch from the Sieve maintainer.
546 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
547 Eximstats - Fixed some broken HTML links and added missing column headers
549 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
550 columns for Rejects, Temp Rejects, Ham, and Spam rows.
552 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
554 PH/43 Yet another patch from the Sieve maintainer.
556 PH/44 I found a way to check for a TCP/IP connection going away before sending
557 the response to the final '.' that terminates a message, but only in the
558 case where the client has not sent further data following the '.'
559 (unfortunately, this is allowed). However, in many cases there won't be
560 any further data because there won't be any more messages to send. A call
561 to select() can be used: if it shows that the input is "ready", there is
562 either input waiting, or the socket has been closed. An attempt to read
563 the next input character can distinguish the two cases. Previously, Exim
564 would have sent an OK response which the client would never have see.
565 This could lead to message repetition. This fix should cure that, at
566 least in a lot of common cases.
568 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
569 advertised in response to EHLO.
575 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
576 fixed by 4.65/MH/01 (is this a record?) are fixed:
578 (i) An empty string was always treated as zero by the numeric comparison
579 operators. This behaviour has been restored.
581 (ii) It is documented that the numeric comparison operators always treat
582 their arguments as decimal numbers. This was broken in that numbers
583 starting with 0 were being interpreted as octal.
585 While fixing these problems I realized that there was another issue that
586 hadn't been noticed. Values of message_size_limit (both the global option
587 and the transport option) were treated as octal if they started with 0.
588 The documentation was vague. These values are now always treated as
589 decimal, and I will make that clear in the documentation.
595 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
596 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
599 MH/01 Don't check that the operands of numeric comparison operators are
600 integers when their expansion is in "skipping" mode (fixes bug
601 introduced by 4.64-PH/07).
603 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
604 child addresses, Exim now panics and dies. Previously, because the count
605 is held in a short int, deliveries were likely to be lost. As such a
606 large number of recipients for a single message is ridiculous
607 (performance will be very, very poor), I have chosen to impose a limit
608 rather than extend the field.
614 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
615 leftover -K file (the existence of which was triggered by #402).
616 While we were at it, introduced process PID as part of the -K
617 filename. This should rule out race conditions when creating
620 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
621 processing considerably. Previous code took too long for large mails,
622 triggering a timeout which in turn triggers #401.
624 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
625 in the DK code in transports.c. sendfile() is not really portable,
626 hence the _LINUX specificness.
628 TF/01 In the add_headers option to the mail command in an Exim filter,
629 there was a bug that Exim would claim a syntax error in any
630 header after the first one which had an odd number of characters
633 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
634 callout verification, Exim cached a "reject" for the entire domain. This
635 is correct for most verifications, but it is not correct for a recipient
636 verification with use_sender or use_postmaster set, because in that case
637 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
638 case of MAIL FROM:<> rejection from other early rejections (e.g.
639 rejection of HELO). When verifying a recipient using a non-null MAIL
640 address, the cache is ignored if it shows MAIL FROM:<> rejection.
641 Whatever the result of the callout, the value of the domain cache is
642 left unchanged (for any other kind of callout, getting as far as trying
643 RCPT means that the domain itself is ok).
645 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
648 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
649 manifest itself as EPIPE rather than ECONNECT. When tidying away a
650 session, the daemon ignores ECONNECT errors and logs others; it now
651 ignores EPIPE as well.
653 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
654 (quoted-printable decoding).
656 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
657 later the small subsequent patch to fix an introduced bug.
659 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
661 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
663 PH/08 An error is now given if message_size_limit is specified negative.
665 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
666 to be given (somewhat) arbitrary names.
668 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
671 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
672 miscellaneous code fixes
674 PH/10 Added the log_reject_target ACL modifier to specify where to log
677 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
678 hostname. This is wrong, because it relates to the incoming message (and
679 probably the interface on which it is arriving) and not to the outgoing
680 callout (which could be using a different interface). This has been
681 changed to use the value of the helo_data option from the smtp transport
682 instead - this is what is used when a message is actually being sent. If
683 there is no remote transport (possible with a router that sets up host
684 addresses), $smtp_active_hostname is used.
686 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
687 tweaks were necessary in order to get it to work (see also 21 below):
688 (a) The code assumed that strncpy() returns a negative number on buffer
689 overflow, which isn't the case. Replaced with Exim's string_format()
691 (b) There were several signed/unsigned issues. I just did the minimum
692 hacking in of casts. There is scope for a larger refactoring.
693 (c) The code used strcasecmp() which is not a standard C function.
694 Replaced with Exim's strcmpic() function.
695 (d) The code set only $1; it now sets $auth1 as well.
696 (e) A simple test gave the error "authentication client didn't specify
697 service in request". It would seem that Dovecot has changed its
698 interface. Fortunately there's a specification; I followed it and
699 changed what the client sends and it appears to be working now.
701 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
704 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
705 address A is aliased to B and C, where B exists and C does not. Without
706 -v the output is "A verified" because verification stops after a
707 successful redirection if more than one address is generated. However,
708 with -v the child addresses are also verified. Exim was outputting "A
709 failed to verify" and then showing the successful verification for C,
710 with its parentage. It now outputs "B failed to verify", showing B's
711 parentage before showing the successful verification of C.
713 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
714 look up a TXT record in a specific list after matching in a combined
717 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
718 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
719 they consult the DNS. I had assumed they would set it the way they
720 wanted; and indeed my experiments on Linux seem to show that in some
721 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
722 To be on the safe side, however, I have now made the interface to
723 host_find_byname() similar to host_find_bydns(), with an argument
724 containing the DNS resolver options. The host_find_byname() function now
725 sets these options at its start, just as host_find_bydns() does. The smtp
726 transport options dns_qualify_single and dns_search_parents are passed to
727 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
728 of host_find_byname() use the default settings of RES_DEFNAMES
729 (qualify_single) but not RES_DNSRCH (search_parents).
731 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
732 spool_read_header() do less string testing, by means of a preliminary
733 switch on the second character of optional "-foo" lines. (This is
734 overdue, caused by the large number of possibilities that now exist.
735 Originally there were few.) While I was there, I also converted the
736 str(n)cmp tests so they don't re-test the leading "-" and the first
737 character, in the hope this might squeeze out yet more improvement.
739 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
740 flag allowing group syntax was set by the header_syntax check but not
741 turned off, possible causing trouble later; (2) The flag was not being
742 set at all for the header_verify test, causing "group"-style headers to
743 be rejected. I have now set it in this case, and also caused header_
744 verify to ignore an empty address taken from a group. While doing this, I
745 came across some other cases where the code for allowing group syntax
746 while scanning a header line wasn't quite right (mostly, not resetting
747 the flag correctly in the right place). These bugs could have caused
748 trouble for malformed header lines. I hope it is now all correct.
750 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
751 with the "reply" argument non-NULL. The code, however (which originally
752 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
753 but it didn't always do it. This confused somebody who was copying the
754 code for some other use. I have removed all the tests.
756 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
757 feature that was used to support insecure browsers during the U.S. crypto
758 embargo. It requires special client support, and Exim is probably the
759 only MTA that supported it -- and would never use it because real RSA is
760 always available. This code has been removed, because it had the bad
761 effect of slowing Exim down by computing (never used) parameters for the
762 RSA_EXPORT functionality.
764 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
765 authenticator to fail if there's a tab character in the incoming data
766 (there should never be unless someone is messing about, as it's supposed
767 to be base64-encoded). Also added, on Timo's advice, the "secured" option
768 if the connection is using TLS or if the remote IP is the same as the
769 local IP, and the "valid-client-cert option" if a client certificate has
772 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
773 authenticators. This can be used for authorization after authentication
774 succeeds. (In the case of plaintext, it servers for both authentication
777 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
778 if any retry times were supplied.
780 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
781 connection, where there is no client IP address to check. In this
782 situation, the verify now always succeeds.
784 PH/25 Applied John Jetmore's -Mset patch.
786 PH/26 Added -bem to be like -Mset, but loading a message from a file.
788 PH/27 In a string expansion for a processed (not raw) header when multiple
789 headers of the same name were present, leading whitespace was being
790 removed from all of them, but trailing whitespace was being removed only
791 from the last one. Now trailing whitespace is removed from each header
792 before concatenation. Completely empty headers in a concatenation (as
795 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
796 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
798 PH/29 [Removed. This was a change that I later backed out, and forgot to
799 correct the ChangeLog entry (that I had efficiently created) before
800 committing the later change.]
802 PH/30 Exim was sometimes attempting to deliver messages that had suffered
803 address errors (4xx response to RCPT) over the same connection as other
804 messages routed to the same hosts. Such deliveries are always "forced",
805 so retry times are not inspected. This resulted in far too many retries
806 for the affected addresses. The effect occurred only when there were more
807 hosts than the hosts_max_try setting in the smtp transport when it had
808 the 4xx errors. Those hosts that it had tried were not added to the list
809 of hosts for which the message was waiting, so if all were tried, there
810 was no problem. Two fixes have been applied:
812 (i) If there are any address or message errors in an SMTP delivery, none
813 of the hosts (tried or untried) are now added to the list of hosts
814 for which the message is waiting, so the message should not be a
815 candidate for sending over the same connection that was used for a
816 successful delivery of some other message. This seems entirely
817 reasonable: after all the message is NOT "waiting for some host".
818 This is so "obvious" that I'm not sure why it wasn't done
819 previously. Hope I haven't missed anything, but it can't do any
820 harm, as the worst effect is to miss an optimization.
822 (ii) If, despite (i), such a delivery is accidentally attempted, the
823 routing retry time is respected, so at least it doesn't keep
824 hammering the server.
826 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
827 in ${readsocket because some servers need this prod.
829 PH/32 Added some extra debug output when updating a wait-xxx database.
831 PH/33 The hint "could be header name not terminated by colon", which has been
832 given for certain expansion errors for a long time, was not being given
833 for the ${if def:h_colon_omitted{... case.
835 PH/34 The spec says: "With one important exception, whenever a domain list is
836 being scanned, $domain contains the subject domain." There was at least
837 one case where this was not true.
839 PH/35 The error "getsockname() failed: connection reset by peer" was being
840 written to the panic log as well as the main log, but it isn't really
841 panic-worthy as it just means the connection died rather early on. I have
842 removed the panic log writing for the ECONNRESET error when getsockname()
845 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
846 runs only) independently of the message's sender address. This meant
847 that, if the 4xx error was in fact related to the sender, a different
848 message to the same recipient with a different sender could confuse
849 things. In particualar, this can happen when sending to a greylisting
850 server, but other circumstances could also provoke similar problems.
851 I have changed the default so that the retry time for these errors is now
852 based a combination of the sender and recipient addresses. This change
853 can be overridden by setting address_retry_include_sender=false in the
856 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
857 remote server are returned as part of bounce messages. This was not
858 happening for LMTP over a pipe (the lmtp transport), but now it is the
859 same for both kinds of LMTP.
861 PH/38 Despite being documented as not happening, Exim was rewriting addresses
862 in header lines that were in fact CNAMEs. This is no longer the case.
864 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
865 and queue runs started by the daemon processed all messages. This has
866 been fixed so that -R and -S can now usefully be given with -q<time>.
868 PH/40 Import PCRE release 6.7 (fixes some bugs).
870 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
872 PH/42 Give an error if -q is specified more than once.
874 PH/43 Renamed the variables $interface_address and $interface_port as
875 $received_ip_address and $received_port, to make it clear that these
876 values apply to message reception, and not to the outgoing interface when
877 a message is delivered. (The old names remain recognized, of course.)
879 PH/44 There was no timeout on the connect() call when using a Unix domain
880 socket in the ${readsocket expansion. There now is.
882 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
883 be meaningful with "accept".
885 SC/01 Eximstats V1.43
886 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
888 SC/02 Eximstats V1.44
889 Use a glob alias rather than an array ref in the generated
890 parser. This improves both readability and performance.
892 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
893 Collect SpamAssassin and rejection statistics.
894 Don't display local sender or destination tables unless
895 there is data to show.
896 Added average volumes into the top table text output.
898 SC/04 Eximstats V1.46
899 Collect data on the number of addresses (recipients)
900 as well as the number of messages.
902 SC/05 Eximstats V1.47
903 Added 'Message too big' to the list of mail rejection
904 reasons (thanks to Marco Gaiarin).
906 SC/06 Eximstats V1.48
907 Mainlog lines which have GMT offsets and are too short to
908 have a flag are now skipped.
910 SC/07 Eximstats V1.49 (Alain Williams)
911 Added the -emptyok flag.
913 SC/08 Eximstats V1.50
914 Fixes for obtaining the IP address from reject messages.
916 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
917 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
918 whitesspace changes from 4.64-PH/27
920 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
923 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
924 are found, allow negative numbers in numeric criteria)
926 JJ/06 exipick.20061117.2, added new $message_body_missing variable
928 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
929 to match changes made in 4.64-PH/43
931 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
933 PH/47 Put in an explicit test for a DNS lookup of an address record where the
934 "domain" is actually an IP address, and force a failure. This locks out
935 those revolvers/nameservers that support "A-for-A" lookups, in
936 contravention of the specifications.
938 PH/48 When a host name was looked up from an IP address, and the subsequent
939 forward lookup of the name timed out, the host name was left in
940 $sender_host_name, contrary to the specification.
942 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
943 restricted to single-key lookups, Exim was not diagnosing an error if
944 * or *@ was used with a query-style lookup.
946 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
948 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
949 long ago, but noone interested enough thought of it. Let's just say that
950 the "1.1" means that there are some new functions that weren't there at
951 some point in the past.
953 PH/51 Error processing for expansion failure of helo_data from an smtp
954 transport during callout processing was broken.
956 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
957 tested/used via the -bh/-bhc/-bs options.
959 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
960 bug, fixed in subsequent PCRE releases).
962 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
963 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
965 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
971 SC/01 Use a glob alias rather than an array ref in eximstats generated
972 parser. This improves both readability and performance.
974 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
975 Don't display local sender or destination tables in eximstats unless
976 there is data to show.
977 Added average volumes into the eximstats top table text output.
979 SC/03 Collect data on the number of addresses (recipients) as well
980 as the number of messages in eximstats.
982 TF/01 Correct an error in the documentation for the redirect router. Exim
983 does (usually) call initgroups() when daemonizing.
985 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
986 with consistent privilege compared to when running as a daemon.
988 TF/03 Note in the spec that $authenticated_id is not set for local
989 submissions from trusted users.
991 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
992 Thanks to Dean Brooks <dean@iglou.com> for the patch.
994 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
995 by adding some example configuration directives to the default
996 configuration file. A little bit of work is required to uncomment the
997 directives and define how usernames and passwords are checked, but
998 there is now a framework to start from.
1000 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1001 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1002 without this. I don't know how relevant this is to other LDAP libraries.
1004 PH/02 Add the verb name to the "unknown ACL verb" error.
1006 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1008 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1010 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1011 directory not expanded when it should be if an expanded home directory
1012 was set for the address (which is overridden by the transport).
1014 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1017 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1018 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1019 because it is too late at that time, and has no effect.
1021 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1022 security issue with \' (bugzilla #107). I could not use the
1023 PQescapeStringConn() function, because it needs a PGconn value as one of
1026 PH/08 When testing addresses using -bt, indicate those final addresses that
1027 are duplicates that would not cause an additional delivery. At least one
1028 person was confused, thinking that -bt output corresponded to deliveries.
1029 (Suppressing duplicates isn't a good idea as you lose the information
1030 about possibly different redirections that led to the duplicates.)
1032 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1033 systems where poll() doesn't work, in particular OS X.
1035 PH/10 Added more information to debugging output for retry time not reached.
1037 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1038 operations in malware.c.
1040 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1043 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1044 syslog_duplication set false, log lines that would normally be written
1045 both the the main log and to the reject log were not written to syslog at
1048 PH/14 In the default configuration, change the use of "message" in ACL warn
1049 statements to "add_header".
1051 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1052 not followed by a command (e.g. "seen endif").
1054 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1055 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1058 PH/17 Added extra conditions to the default value of delay_warning_condition
1062 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1063 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1064 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1067 The Auto-Submitted: and various List- headers are standardised, whereas I
1068 don't think Precedence: ever was.
1070 PH/18 Refactored debugging code in route_finduser() to show more information,
1071 in particular, the error code if getpwnam() issues one.
1073 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1074 This is apparently needed in addition to the PH/07 change above to avoid
1075 any possible encoding problems.
1077 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1078 but not after initializing Perl.
1080 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1081 output them only if debugging. By default they are written stderr,
1082 apparently, which is not desirable.
1084 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1087 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1090 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1092 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1093 authenticated or an ident call has been made. Suppress the default
1094 values for $authenticated_id and $authenticated_sender (but permit -oMai
1095 and -oMas) when testing with -bh.
1097 PH/24 Re-jigged the order of the tests in the default configuration so that the
1098 tests for valid domains and recipients precede the DNS black list and CSA
1099 tests, on the grounds that those ones are more expensive.
1101 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1102 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1103 This bug exists in every version of Exim that I still have, right back to
1106 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1107 However, an attempt to turn on case-sensitivity in a regex key by
1108 including (?-i) didn't work because the subject string was already
1109 lowercased, and the effects were non-intuitive. It turns out that a
1110 one-line patch can be used to allow (?-i) to work as expected.
1116 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1117 other effects) broke the use of negated acl sub-conditions.
1119 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1122 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1123 "Deny" causes Exim to reject the incoming connection with a 554 error.
1124 Unfortunately, if there is a major crisis, such as a disk failure,
1125 tcp-wrappers gives "deny", whereas what one would like would be some
1126 kind of temporary error. A kludge has been added to help with this.
1127 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1128 554 error is used if errno is still zero or contains ENOENT (which occurs
1129 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1132 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1134 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1135 errors. Otherwise a message that provokes a temporary error (when other
1136 messages do not) can cause a whole host to time out.
1138 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1139 addresses were routed directly to files or pipes from a redirect router.
1140 File deliveries just didn't batch; pipe deliveries might have suffered
1143 PH/06 A failure to get a lock for a hints database would erroneously always say
1144 "Failed to get write lock", even when it was really a read lock.
1146 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1147 of 0600. This has been changed to use the value of the lockfile_mode
1148 option (which defaults to 0600).
1150 PH/08 Applied small patch from the Sieve maintainer.
1152 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1153 folder from quota calculations, a direct delivery into this folder messed
1154 up the contents of the maildirsize file. This was because the regex was
1155 used only to exclude .Trash (or whatever) when the size of the mailbox
1156 was calculated. There was no check that a delivery was happening into an
1157 excluded directory. This bug has been fixed by ignoring all quota
1158 processing for deliveries into excluded directories.
1160 PH/10 Added the maildirfolder_create_regex option to appendfile.
1166 PH/01 The code for finding all the local interface addresses on a FreeBSD
1167 system running IPv6 was broken. This may well have applied to all BSD
1168 systems, as well as to others that have similar system calls. The broken
1169 code found IPv4 interfaces correctly, but gave incorrect values for the
1170 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1171 that it would not match correctly against @[] and not recognize the IPv6
1174 PH/02 The ipliteral router was not recognizing addresses of the form user@
1175 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1177 PH/03 Added disable_ipv6.
1179 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1180 decoded form, because it is most often used to construct To: headers
1181 lines in autoreplies, and the decoded form may well be syntactically
1182 invalid. However, $reply_address has leading white space removed, and all
1183 newlines turned into spaces so that the autoreply transport does not
1186 PH/05 If group was specified without a user on a router, and no group or user
1187 was specified on a transport, the group from the router was ignored.
1189 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1190 for visible compile-time settings that can be used to change these
1191 numbers, for those that want even more. Backwards compatibility with old
1192 spool files has been maintained. However, going back to a previous Exim
1193 release will lost any variables that are in spool files.
1195 PH/07 Two small changes when running in the test harness: increase delay when
1196 passing a TCP/IP connection to a new process, in case the original
1197 process has to generate a bounce, and remove special handling of
1198 127.0.0.2 (sic), which is no longer necessary.
1200 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1201 be the same on different OS.
1203 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1206 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1207 whether --show-vars was specified or not
1209 JJ/02 exipick: Added support for new ACL variable spool format introduced
1212 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1213 syntactically invalid From: or Reply-to: line, and a filter used this to
1214 generate an autoreply, and therefore failed to obtain an address for the
1215 autoreply, Exim could try to deliver to a non-existent relative file
1216 name, causing unrelated and misleading errors. What now happens is that
1217 it logs this as a hard delivery error, but does not attempt to create a
1220 PH/11 The exinext utility has a -C option for testing purposes, but although
1221 the given file was scanned by exinext itself; it wasn't being passed on
1222 when Exim was called.
1224 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1225 an end-of-file indication when reading a command response.
1227 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1228 compiled. In many other places in Exim, IPv6 addresses are always
1229 recognized, so I have changed this. It also means that IPv4 domain
1230 literals of the form [IPV4:n.n.n.n] are now always recognized.
1232 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1233 used if the router is not running as root, for example, when verifying at
1234 ACL time, or when using -bh. The debugging output from this situation was
1235 non-existent - all you got was a failure to exec. I have made two
1238 (a) Failures to set uid/gid, the current directory, or a process leader
1239 in a subprocess such as that created by queryprogram now generate
1240 suitable debugging ouput when -d is set.
1242 (b) The queryprogram router detects when it is not running as root,
1243 outputs suitable debugging information if -d is set, and then runs
1244 the subprocess without attempting to change uid/gid.
1246 PH/15 Minor change to Makefile for building test_host (undocumented testing
1249 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1250 additional section of a DNS packet that returns MX or SRV records.
1251 Instead, it always explicitly searches for A/AAAA records. This avoids
1252 major problems that occur when a DNS server includes only records of one
1253 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1254 fixed another bug: if SRV records were looked up and the corresponding
1255 address records were *not* found in the additional section, the port
1256 values from the SRV records were lost.
1258 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1259 using the correct key (the original address) when searching the retry
1260 rules in order to find which one to use for generating the retry hint.
1262 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1263 adding the default one. Similarly, if it contains a Reply-To: header, the
1264 errors_reply_to option, if set, is not used.
1266 PH/19 When calculating a retry time, Exim used to measure the "time since
1267 failure" by looking at the "first failed" field in the retry record. Now
1268 it does not use this if it is later than than the arrival time of the
1269 message. Instead it uses the arrival time. This makes for better
1270 behaviour in cases where some deliveries succeed, thus re-setting the
1271 "first failed" field. An example is a quota failure for a huge message
1272 when small messages continue to be delivered. Without this change, the
1273 "time since failure" will always be short, possible causing more frequent
1274 delivery attempts for the huge message than are intended.
1275 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1277 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1278 $1, $2, $3) because the numerical variables can be reset during some
1279 expansion items (e.g. "match"), thereby losing the authentication data.
1281 PH/21 Make -bV show the size of off_t variables so that the test suite can
1282 decide whether to run tests for quotas > 2G.
1284 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1285 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1286 filecount value is greater than 2G or if a quota value is greater than 2G
1287 on a system where the size of off_t is not greater than 4, a panic error
1290 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1291 never match. The debug and -bh output now contains an explicit error
1292 message indicating a malformed IPv4 address or mask.
1294 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1295 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1296 PH/23 above applies.
1298 PH/25 Do not write to syslog when running in the test harness. The only
1299 occasion when this arises is a failure to open the main or panic logs
1300 (for which there is an explicit test).
1302 PH/26 Added the /no_tell option to "control=freeze".
1304 PH/27 If a host name lookup failed very early in a connection, for example, if
1305 the IP address matched host_lookup and the reverse lookup yielded a name
1306 that did not have a forward lookup, an error message of the form "no IP
1307 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1308 could be logged. Now it outputs the IP address instead of "NULL".
1310 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1311 allows the sender and the authenticated sender to be set when
1312 submitting a message from within Exim. Since child_open_exim() is
1313 documented for local_scan(), the new function should be too.
1315 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1316 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1317 results in an empty string is now treated as unset.
1319 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1321 PH/31 Added sender_verify_fail logging option.
1323 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1324 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1325 tidied the source and removed it altogether.
1327 PH/33 When a queue run was abandoned because the load average was too high, a
1328 log line was always written; now it is written only if the queue_run log
1329 selector is set. In addition, the log line for abandonment now contains
1330 information about the queue run such as the pid. This is always present
1331 in "start" and "stop" lines but was omitted from the "abandon" line.
1333 PH/34 Omit spaces between a header name and the colon in the error message that
1334 is given when verify = headers_syntax fails (if there are lots of them,
1335 the message gets confusing).
1337 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1338 names, as there are now some PTR records that contain slashes. This check
1339 is only to protect against broken name servers that fall over on strange
1340 characters, so the fact that it applies to all lookups doesn't matter.
1342 PH/36 Now that the new test suite is complete, we can remove some of the
1343 special code in Exim that was needed for the old test suite. For example,
1344 sorting DNS records because real resolvers return them in an arbitrary
1345 order. The new test suite's fake resolver always returns records in the
1348 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1349 bounces) except when queue_only is set, to avoid logging races between
1350 the different processes.
1352 PH/38 Panic-die if .include specifies a non-absolute path.
1354 PH/39 A tweak to the "H" retry rule from its user.
1356 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1357 a label. They prevented compilation on older perls.
1359 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1360 a warning to be raised on newish perls.
1362 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1363 on queue. Changes to match documented behaviour of showing count of
1364 messages matching specified criteria.
1366 PH/40 Changed the default ident timeout from 30s to 5s.
1368 PH/41 Added support for the use of login_cap features, on those BSD systems
1369 that have them, for controlling the resources used by pipe deliveries.
1371 PH/42 The content-scanning code uses fopen() to create files in which to put
1372 message data. Previously it was not paying any attention to the mode of
1373 the files. Exim runs with umask(0) because the rest of the code creates
1374 files with open(), and sets the required mode explicitly. Thus, these
1375 files were ending up world-writeable. This was not a big issue, because,
1376 being within the spool directory, they were not world-accessible. I have
1377 created a function called modefopen, which takes an additional mode
1378 argument. It sets umask(777), creates the file, chmods it to the required
1379 mode, then resets the umask. All the relevant calls to fopen() in the
1380 content scanning code have been changed to use this function.
1382 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1383 to 24 hours. This avoids potential overflow problems when processing G
1384 and H retry rules. I suspect nobody ever tinkers with this value.
1386 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1388 PH/45 When the plaintext authenticator is running as a client, the server's
1389 challenges are checked to ensure they are valid base64 strings. By
1390 default, the authentication attempt is cancelled if an invalid string is
1391 received. Setting client_ignore_invalid_base64 true ignores these errors.
1392 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1393 they are received. Thus, the responses can be made to depend on the
1394 challenges. If an invalid string is ignored, an empty string is placed in
1397 PH/46 Messages that are created by the autoreply transport now contains a
1398 References: header, in accordance with RFCs 2822 and 3834.
1400 PH/47 Added authenticated_sender_force to the smtp transport.
1402 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1404 PH/49 Installed latest patch from the Sieve maintainer.
1406 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1407 was also set, the appendfile transport was unnecessarily scanning a
1408 directory of message files (e.g. for maildir delivery) to find the count
1409 of files (along with the size), even though it did not need this
1410 information. It now does the scan only if it needs to find either the
1411 size of the count of files.
1413 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1415 PH/52 Two bugs concerned with error handling when the smtp transport is
1418 (i) Exim was not creating retry information for temporary errors given
1419 for individual recipients after the DATA command when the smtp transport
1420 was used in LMTP mode. This meant that they could be retried too
1421 frequently, and not timed out correctly.
1423 (ii) Exim was setting the flag that allows error details to be returned
1424 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1425 recipients that were returned after the DATA command.
1427 PH/53 This is related to PH/52, but is more general: for any failing address,
1428 when detailed error information was permitted to be returned to the
1429 sender, but the error was temporary, then after the final timeout, only
1430 "retry timeout exceeded" was returned. Now it returns the full error as
1431 well as "retry timeout exceeded".
1433 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1434 do this, and (what is worse) MTAs that accept it.
1436 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1437 will now be deprecated.
1439 PH/56 New os.c-cygwin from the Cygwin maintainer.
1441 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1442 formats (previously only available in exim formats via -bpr, -bpru,
1443 and -bpra. Now also available in native and exiqgrep formats)
1445 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1446 with very large, slow to parse queues
1448 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1450 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1452 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1453 responses to authentication challenges, though it was showing the
1454 challenges; (ii) I've removed the CR characters from the debug output for
1457 PH/58 Allow for the insertion of a newline as well as a space when a string
1458 is turned into more than one encoded-word during RFC 2047 encoding. The
1459 Sieve code now uses this.
1461 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1462 data_4xx, lost_connection, tls_required.
1464 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1465 message was being sent as an SMTP response.
1467 PH/61 Add -l and -k options to exicyclog.
1469 PH/62 When verifying, if an address was redirected to one new address, so that
1470 verification continued, and the new address failed or deferred after
1471 having set something in $address_data, the value of $address_data was not
1472 passed back to the ACL. This was different to the case when no
1473 redirection occurred. The value is now passed back in both cases.
1475 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1476 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1477 use login_cap.h, so on its own it isn't the distinguishing feature. The
1478 new name refers directly to the setclassresources() function.
1480 PH/65 Added configuration files for NetBSD3.
1482 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1484 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1485 is preferred over IPv4.
1487 PH/68 The bounce_return_message and bounce_return_body options were not being
1488 honoured for bounces generated during the reception of non-SMTP messages.
1489 In particular, this applied to messages rejected by the ACL. This bug has
1490 been fixed. However, if bounce_return_message is true and bounce_return_
1491 body is false, the headers that are returned for a non-SMTP message
1492 include only those that have been read before the error was detected.
1493 (In the case of an ACL rejection, they have all been read.)
1495 PH/69 The HTML version of the specification is now built in a directory called
1496 spec_html instead of spec.html, because the latter looks like a path with
1497 a MIME-type, and this confuses some software.
1499 PH/70 Catch two compiler warnings in sieve.c.
1501 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1502 function verify_get_ident() calls ip_connect() to connect a socket, but
1503 if the "connect()" function timed out, ip_connect() used to close the
1504 socket. However, verify_get_ident() also closes the socket later, and in
1505 between Exim writes to the log, which may get opened at this point. When
1506 the socket was closed in ip_connect(), the log could get the same file
1507 descriptor number as the socket. This naturally causes chaos. The fix is
1508 not to close the socket in ip_connect(); the socket should be closed by
1509 the function that creates it. There was only one place in the code where
1510 this was missing, in the iplookup router, which I don't think anybody now
1511 uses, but I've fixed it anyway.
1513 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1514 well as to direct DNS lookups. Otherwise the handling of names in host
1515 lists is inconsistent and therefore confusing.
1521 PH/01 Two changes to the default runtime configuration:
1523 (1) Move the checks for relay_from_hosts and authenticated clients from
1524 after to before the (commented out) DNS black list checks.
1526 (2) Add control=submission to the relay_from_hosts and authenticated
1527 clients checks, on the grounds that messages accepted by these
1528 statements are most likely to be submissions.
1530 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1532 (1) Generate an error if the third argument for the ${prvs expansion is
1535 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1538 (3) Reset the variables that are obtained from the first argument of
1539 ${prvscheck and used in the second argument before leaving the code,
1540 because their memory is reclaimed, so using them afterwards may do
1543 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1544 one (it's much easier than Tom thought :-).
1546 (5) Because of (4), we can now allow for the use of $prvscheck_result
1547 inside the third argument.
1549 PH/03 For some reason, the default setting of PATH when running a command from
1550 a pipe transport was just "/usr/bin". I have changed it to
1553 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1554 anything to be listed in the output from -bV.
1556 PH/05 When a filter generated an autoreply, the entire To: header line was
1557 quoted in the delivery log line, like this:
1559 => >A.N.Other <ano@some.domain> <original@ddress> ...
1561 This has been changed so that it extracts the operative address. There
1562 may be more than one such address. If so, they are comma-separated, like
1565 => >ano@some.domain,ona@other.domain <original@ddress> ...
1567 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1568 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1569 not being looked up in the rDNS to get a host name, Exim was showing the
1570 IP address twice in Received: lines, even though the IP addresses were
1571 identical. For example:
1573 Received: from [1.2.3.4] (helo=[1.2.3.4])
1575 However, if the real host name was known, it was omitting the HELO data
1576 if it matched the actual IP address. This has been tidied up so that it
1577 doesn't show the same IP address twice.
1579 PH/07 When both +timestamp and +memory debugging was on, the value given by
1580 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1581 called by the debug printing, thereby overwriting the timestamp buffer.
1582 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1584 PH/08 When the original message was included in an autoreply transport, it
1585 always said "this is a copy of the message, including all the headers",
1586 even if body_only or headers_only was set. It now gives an appropriate
1589 PH/09 Applied a patch from the Sieve maintainer which:
1591 o fixes some comments
1592 o adds the (disabled) notify extension core
1593 o adds some debug output for the result of if/elsif tests
1594 o points to the current vacation draft in the documentation
1595 and documents the missing references header update
1599 o fixes a bug in processing the envelope test (when testing
1600 multiple envelope elements, the last element determinted the
1603 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1604 Electronic Mail") by including:
1606 Auto-submitted: auto-generated
1608 in the messages that it generates (bounce messages and others, such as
1609 warnings). In the case of bounce messages for non-SMTP mesages, there was
1610 also a typo: it was using "Auto_submitted" (underscore instead of
1611 hyphen). Since every message generated by Exim is necessarily in response
1612 to another message, thes have all been changed to:
1614 Auto-Submitted: auto-replied
1616 in accordance with these statements in the RFC:
1618 The auto-replied keyword:
1620 - SHOULD be used on messages sent in direct response to another
1621 message by an automatic process,
1623 - MUST NOT be used on manually-generated messages,
1625 - MAY be used on Delivery Status Notifications (DSNs) and Message
1626 Disposition Notifications (MDNs),
1628 - MUST NOT be used on messages generated by automatic or periodic
1629 processes, except for messages which are automatic responses to
1632 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1633 to the default Received: header definition.
1635 PH/12 Added log selector acl_warn_skipped (default on).
1637 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1638 variables because (a) they are in the wrong storage pool and (b) even if
1639 they were copied, it wouldn't work properly because of the caching.
1641 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1642 checking when decoding. Apparently there are clients that generate
1643 overlong encoded strings. Why am I not surprised?
1645 PH/15 If the first argument of "${if match_address" was not empty, but did not
1646 contain an "@" character, Exim crashed. Now it writes a panic log message
1647 and treats the condition as false.
1649 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1651 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1652 extension "envelope-auth". The code is finished and in agreement with
1653 other implementations, but there is no documentation so far and in fact,
1654 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1655 not changing the active code.
1657 Print executed "if" and "elsif" statements when debugging is used. This
1658 helps a great deal to understand what a filter does.
1660 Document more things not specified clearly in RFC3028. I had all this
1661 sorted out, when out of a sudden new issues came to my mind. Oops."
1663 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1666 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1667 canonical form (as documented). However, after a host name lookup from
1668 the IP address, check_host() was doing a simple string comparison with
1669 addresses acquired from the DNS when checking that the found name did
1670 have the original IP as one of its addresses. Since any found IPv6
1671 addresses are likely to be in abbreviated form, the comparison could
1672 fail. Luckily, there already exists a function for doing the comparison
1673 by converting both addresses to binary, so now that is used instead of
1674 the text comparison.
1676 PH/20 There was another similar case to PH/19, when a complete host name was
1677 given in a host list; looking up its IP address could give an abbreviated
1678 form, whereas the current host's name might or might not be abbreviated.
1679 The same fix has been applied.
1685 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1686 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1689 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1690 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1692 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1694 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1695 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1696 Also, for "Auto-Submitted", treat anything other than "no" as
1697 non-personal, in accordance with RFC 3834. (Previously it treated
1698 anything starting "auto-" as non-personal.)
1700 TF/01 The control=submission/name=... option had a problem with syntax
1701 errors if the name included a slash character. The /name= option
1702 now slurps the rest of the string, so it can include any characters
1703 but it must come last in the list of options (after /sender_retain
1706 PH/05 Some modifications to the interface to the fake nameserver for the new
1714 TK/01 Added the "success_on_redirect" address verification option. See
1715 NewStuff for rationale and an example.
1717 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1719 PH/02 Patch to exigrep to allow it to work on syslog lines.
1721 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1722 fread() to skip over the body file's header line, because in Cygwin the
1723 header line is locked and is inaccessible.
1725 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1726 co-exist for some time) to make it clear that it is the Exim ID that is
1727 referenced, not the Message-ID: header line.
1729 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1730 string_format() function, because snprintf() does not exist on all
1733 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1734 ${stat: expansion item.
1736 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1737 protocol synchronization error", to keep the pedants happy.
1739 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1740 well as for IRIX systems, when gcc is being used. See the host.c source
1743 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1745 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1748 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1749 transport and to the smtp transport in LMTP mode.
1751 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1753 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1754 run a filter in a subprocess. This could lead to confusion in subsequent
1755 lookups in the parent process. There should also be a search_tidyup() at
1756 the end of the subprocess.
1758 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1759 only if the host matched helo_try_verify_hosts, which caused the
1760 verification to occur when the EHLO/HELO command was issued. The ACL just
1761 tested the remembered result. Now, if a previous verification attempt has
1762 not happened, "verify = helo" does it there and then.
1764 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1766 TK/03 Fix log output including CR from clamd.
1768 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1769 exist provoked a memory error which could cause a segfault.
1771 PH/15 Installed PCRE 6.2
1773 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1775 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1776 of the problem. Specifically, suggested +O2 rather than +O1 for the
1779 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1781 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1782 record, the port was not correctly logged when the outgoing_port log
1783 selector was set (it logged the transort's default port).
1785 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1786 fallback_hosts, and "hosts" in the smtp transport.
1788 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1789 host errors such as "Connection refused".
1791 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1792 authentication with radiusclient 0.4.9:
1794 - Error returned from rc_read_config was caught wrongly
1795 - Username/password not passed on to radius server due to wrong length.
1797 The presumption is that some radiusclient API changes for 4.51/PH/17
1798 were not taken care of correctly. The code is still untested by me (my
1799 Linux distribution still has 0.3.2 of radiusclient), but it was
1800 contributed by a Radius user.
1802 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1803 expanding the "port" option of the smtp transport.
1805 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1806 while reading a MIME header. Thanks to Tom Hughes for a patch.
1808 PH/24 Include config.h inside local_scan.h so that configuration settings are
1811 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1812 that in an ACL for RCPT (for example), you can examine exactly what was
1815 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1816 commands, but it was not correctly comparing the address with the actual
1817 client host address. Thus, it would show the EHLO address in Received:
1818 header lines when this was not necessary.
1820 PH/27 Added the % operator to ${eval:}.
1822 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1823 it should be ignoring failures (because with -C, for example, it has lost
1824 privilege). It wasn't ignoring creation failures other than "already
1827 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1830 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1831 because an input line was too long, either on its own, or by virtue of
1832 too many continuations, the temporary file was not being removed, and the
1833 return code was incorrect.
1835 PH/31 Missing "BOOL" in function definition in filtertest.c.
1837 PH/32 Applied Sieve patches from the maintainer.
1839 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1841 PH/33 Added "verify = not_blind".
1843 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1844 Local/Makefile (with some defaults set). These are used in built scripts
1845 such as exicyclog, but they have never been used in the exim_install
1846 script (though there are many overriding facilities there). I have
1847 arranged that the exim_install script now takes note of these two
1850 PH/35 Installed configuration files for Dragonfly.
1852 PH/36 When a locally submitted message by a trusted user did not contain a
1853 From: header, and the sender address was obtained from -f or from an SMTP
1854 MAIL command, and the trusted user did not use -F to supply a sender
1855 name, $originator_name was incorrectly used when constructing a From:
1856 header. Furthermore, $originator_name was used for submission mode
1857 messages from external hosts without From: headers in a similar way,
1858 which is clearly wrong.
1860 PH/37 Added control=suppress_local_fixups.
1862 PH/38 When log_selector = +received_sender was set, and the addition of the
1863 sender made the log line's construction buffer exactly full, or one byte
1864 less than full, an overflow happened when the terminating "\n" was
1867 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1868 when the result of a list match is failure because a DNS lookup failed.
1870 PH/40 RM_COMMAND is now used in the building process.
1872 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1873 the "build-* directories that it finds.
1875 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1876 address in a domain literal was a prefix of an interface address.
1878 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1879 when verifying a sender address, unless rewrite_headers is false.
1881 PH/44 Wrote a long comment about why errors_to addresses are verified as
1882 recipients, not senders.
1884 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1885 the ratelimit ACL was added.
1887 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1889 PH/46 Added extra information about PostgreSQL errors to the error string.
1891 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1892 avoiding the need to install special zones in a real server. This is
1893 backwards compatible; if it can't find the fake resolver, it drops back.
1894 Thus, both old and new test suites can be run.
1896 TF/02 Added util/ratelimit.pl
1898 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1899 clock is set back in time.
1901 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1902 Candler <B.Candler@pobox.com>.
1904 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1905 OK for addresses that are the result of redirections.
1907 PH/48 A number of further additions for the benefit of the new test suite,
1908 including a fake gethostbyname() that interfaces to the fake DNS resolver
1911 TF/06 The fix for widen_domains has also been applied to qualify_single and
1912 search_parents which are the other dnslookup options that can cause
1915 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1918 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1920 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1921 service" if it encountered an MX record with an empty target hostname.
1922 The message is now "an MX or SRV record indicated no SMTP service".
1924 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1925 if the DNS of the sending site is misconfigured. This is quite a
1926 common situation. This change restores the behaviour of treating a
1927 helo verification defer as a failure.
1929 PH/51 If self=fail was set on a router, the bounce message did not include the
1930 actual error message.
1936 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1938 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1939 command itself ended in error, the underlying message about the transport
1940 filter timeout was being overwritten with the pipe command error. Now the
1941 underlying error message should be appended to the second error message.
1943 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1945 PH/02 Reduce the amount of output that "make" produces by default. Full output
1946 can still be requested.
1948 PH/03 The warning log line about a condition test deferring for a "warn" verb
1949 was being output only once per connection, rather than after each
1950 occurrence (because it was using the same function as for successful
1951 "warn" verbs). This seems wrong, so I have changed it.
1953 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1954 it should not have, which might have caused a crash in the right
1955 circumstances, but probably never did.
1957 PH/04 Installed a modified version of Tony Finch's patch to make submission
1958 mode fix the return path as well as the Sender: header line, and to
1959 add a /name= option so that you can make the user's friendly name appear
1962 TF/03 Added the control = fakedefer ACL modifier.
1964 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1965 Mark Lowes for thorough testing.
1967 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1969 TK/03 Merged latest SRS patch from Miles Wilton.
1971 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1972 with the definition in sysexits.h (which is #included earlier).
1973 Fortunately, Exim does not actually use EX_OK. The code used to try to
1974 preserve the sysexits.h value, by assumimg that macro definitions were
1975 scanned for macro replacements. I have been disabused of this notion,
1976 so now the code just undefines EX_OK before #including unistd.h.
1978 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1979 in the smtp transport. When a block could not be written in a single
1980 write() function, the timeout was being re-applied to each part-write.
1981 This seems wrong - if the receiver was accepting one byte at a time it
1982 would take for ever. The timeout is now adjusted when this happens. It
1983 doesn't have to be particularly precise.
1985 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1986 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1988 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1989 without a domain if the check to <postmaster@domain> fails.
1991 SC/01 Eximstats: added -xls and the ability to specify output files
1992 (patch written by Frank Heydlauf).
1994 SC/02 Eximstats: use FileHandles for outputing results.
1996 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1998 SC/04 Eximstats: fixed display of large numbers with -nvr option
2000 SC/05 Eximstats: fixed merging of reports with empty tables.
2002 SC/06 Eximstats: added the -include_original_destination flag
2004 SC/07 Eximstats: removed tabs and trailing whitespace.
2006 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2008 TK/06 MBOX spool code: Add real "From " MBOX separator line
2009 so the .eml file is really in mbox format (even though
2010 most programs do not really care). Patch from Alex Miller.
2012 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2013 The latter is generated from $received_to and is only set if the
2014 message has one envelope recipient. SA can use these headers,
2015 obviously out-of-the-box. Patch from Alex Miller.
2017 PH/08 The ${def test on a variable was returning false if the variable's
2018 value was "0", contrary to what the specification has always said!
2019 The result should be true unless the variable is empty.
2021 PH/09 The syntax error of a character other than { following "${if
2022 def:variable_name" (after optional whitespace) was not being diagnosed.
2023 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2024 accidental colon was present, for example, could give incorrect results.
2026 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2027 result is used (not including appendfile, where other changes are about
2030 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2031 This involved changing a lot of size variables from int to off_t. It
2032 should work with maildirs and everything.
2034 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2035 spamd dying while we are connected to it.
2037 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2040 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2041 function for BDB 4.3.
2043 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2045 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2046 the sources for PCRE have been re-arranged and more files are now
2049 PH/15 The code I had for printing potentially long long variables in PH/11
2050 above was not the best (it lost precision). The length of off_t variables
2051 is now inspected at build time, and an appropriate printing format (%ld
2052 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2053 to be "long long int" or "long int". This is needed for the internal
2054 formatting function string_vformat().
2056 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2057 the configuration file to be ":syslog", then the script "guesses" where
2058 the logs files are, rather than using the compiled in default. In our
2059 case the guess is not the same as the compiled default, so the script
2060 suddenly stopped working when I started to use syslog. The patch checks
2061 to see if log_file_path is "". If so, it attempts to read it from exim
2062 with no configuration file to get the compiled in version, before it
2063 falls back to the previous guessing code."
2065 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2066 implementing BATV in an Exim configuration. See NewStuff for the gory
2069 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2070 Makefile that are specific to HP-UX.
2072 PH/18 If the "use_postmaster" option was set for a recipient callout together
2073 with the "random" option, the postmaster address was used as the MAIL
2074 FROM address for the random test, but not for the subsequent recipient
2075 test. It is now used for both.
2077 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2078 patch removes a few documentation additions to RFC 3028, because the
2079 latest draft now contains them. It adds the new en;ascii-case comparator
2080 and a new error check for 8bit text in MIME parts. Comparator and
2081 require names are now matched exactly. I enabled the subaddress
2082 extension, but it is not well tested yet (read: it works for me)."
2084 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2085 rework some of the code of TK/09 above to avoid the hardwired use of
2086 "%lld" and "long long". Replaced the call to snprintf() with a call to
2089 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2090 records point to non-existent hosts", "retry timeout exceeded", and
2091 "retry time not reached for any host after a long failure period".
2093 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2094 experimental DomainKeys support:
2096 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2097 (2) On an error such as an illegally used "control", the wrong name for
2098 the control was given.
2100 These problems did NOT occur unless DomainKeys support was compiled.
2102 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2104 PH/24 Added ${if match_ip condition.
2106 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2107 databases so that it will be absolutely obvious if a crash occurs in the
2108 DB library. This is a regular occurrence (often caused by mis-matched
2111 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2112 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2113 picked up on a user's system that detects such things. There doesn't seem
2114 to be a gcc warning option for this - only an attribute that has to be
2115 put on the function's prototype. It seems that in Fedora Core 4 they have
2116 set this on a number of new functions. No doubt there will be more in due
2119 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2120 specify a transport. However, if an address that was verified by such a
2121 router was the subject of a callout, Exim crashed because it tried to
2122 read the rcpt_include_affixes from the non-existent transport. Now it
2123 just assumes that the setting of that option is false. This bug was
2124 introduced by 4.51/PH/31.
2126 PH/28 Changed -d+all to exclude +memory, because that information is very
2127 rarely of interest, but it makes the output a lot bigger. People tend to
2128 do -d+all out of habit.
2130 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2131 code in os-type was giving problems when libc.so lives in lib64, like on
2134 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2135 aren't the modern standard, and it seems that some systems' include files
2136 don't always have them. Exim was already checking for some of the newer
2137 ones like T_AAAA, and defining it itself. I've added checks for all the
2138 record types that Exim uses.
2140 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2141 not automatically generating a new one, as it is supposed to. This
2142 prevented TLS from working. If the file did exist, but contained invalid
2143 data, a new version was generated, as expected. It was only the case of a
2144 non-existent file that was broken.
2146 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2147 with a change in libdomainkeys > 0.64.
2149 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2150 from DNS. If the selector record carries the flag, it now has
2151 precedence over the domain-wide flag.
2153 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2155 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2156 the use of a port name that isn't defined in /etc/services) occurred, the
2157 message was deferred as in a normal delivery, and thus remained on the
2158 spool, instead of being failed because of the mua_wrapper setting. This
2159 is now fixed, and I tidied up some of the mua_wrapper messages at the
2162 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2163 the messages in a hash of arrays rather than using individual hashes.
2164 This is a bit cleaner and results in dramatic memory savings, albeit
2165 at a slight CPU cost.
2167 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2168 as requested by Marc Sherman.
2170 SC/10 Eximstats: added histograms for user specified patterns as requested
2173 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2175 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2176 fopen() in the content-scanning modules that did not already have it.
2182 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2183 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2185 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2187 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2189 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2190 to test Sieve filters that use "vacation".
2192 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2193 that changes the way the GnuTLS parameters are stored in the cache file.
2194 The new format can be generated externally. For backward compatibility,
2195 if the data in the cache doesn't make sense, Exim assumes it has read an
2196 old-format file, and it generates new data and writes a new file. This
2197 means that you can't go back to an older release without removing the
2200 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2201 work if there are any delivery delays because "one_time" forces the
2202 parent to be marked "delivered", so its unseen clone is never tried
2203 again. For this reason, Exim now forbids the simultaneous setting of
2206 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2207 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2208 entry if you want to know the details. The fix, however, neglected to
2209 consider the case when local delivery batching is involved. The test for
2210 "previously delivered" was not happening when checking to see if an
2211 address could be batched with a previous (undelivered) one; under
2212 certain circumstances this could lead to multiple deliveries to the same
2215 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2216 in its include files, and this causes problems building Exim.
2218 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2219 header_syntax) but Exim was just ignoring anything given after a slash.
2220 In particular, this caused confusion with an attempt to use "verify =
2221 reverse_host_lookup/defer_ok". An error is now given when options are
2222 supplied for verify items that do not have them. (Maybe reverse_host_
2223 lookup should have a defer_ok option, but that's a different point.)
2225 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2226 defined by RFC 821) to 2048, because there were problems with some AUTH
2227 commands, and RFC 1869 says the size should be increased for extended
2228 SMTP commands that take arguments.
2230 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2233 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2234 "unknown" error; now it says that the functionality isn't in the binary.
2236 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2237 an address' error message when a string expansion fails (syntax or
2238 whatever). Otherwise the password may appear in the log. Following change
2239 PH/42 below, there is no longer a chance of it appearing in a bounce
2242 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2244 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2245 of its addresses was ever tried. (Bugzilla bug #2.)
2247 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2248 the result incorrectly in the debug output. (It correctly added a newline
2249 to what was transported.)
2251 TF/01 Added $received_time.
2253 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2254 commented out examples of how to interface to a virus scanner and to
2255 SpamAssassin. Also added commented examples of av_scanner and
2256 spamd_address settings.
2258 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2259 and controls are allowed in which ACLs. There were a couple of minor
2260 errors. Some of the entries in the conditions table (which is a table of
2261 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2262 negation of where the condition IS allowed.
2264 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2266 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2267 header file does not have a version number, so I've had to invent a new
2268 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2269 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2270 radiusclient), but it was contributed by a Radius user.
2272 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2273 files or directories, for OpenSSL.
2275 PH/19 When an Exim process that is running as root has to create an Exim log
2276 file, it does so in a subprocess that runs as exim:exim so as to get the
2277 ownership right at creation (otherwise, other Exim processes might see
2278 the file with the wrong ownership). There was no test for failure of this
2279 fork() call, which would lead to the process getting stuck as it waited
2280 for a non-existent subprocess. Forks do occasionally fail when resources
2281 run out. I reviewed all the other calls to fork(); they all seem to check
2284 PH/20 When checking for unexpected SMTP input at connect time (before writing
2285 the banner), Exim was not dealing correctly with a non-positive return
2286 from the read() function. If the client had disconnected by this time,
2287 the result was a log entry for a synchronization error with an empty
2288 string after "input=" when read() returned zero. If read() returned -1
2289 (an event I could not check), uninitialized data bytes were printed.
2290 There were reports of junk text (parts of files, etc) appearing after
2293 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2295 PH/22 Added support for macro redefinition, and (re)definition in between
2296 driver and ACL definitions.
2298 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2299 forgetting to use the resulting value; it was using the unexpanded value.
2301 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2302 hadn't been configured. The fix is from Juergen Kreileder, who
2303 understands it better than I do:
2305 "Here's what I see happening with three configured cyrus_sasl
2306 authenticators configured (plain, login, cram-md5):
2308 On startup auth_cyrus_sasl_init() gets called for each of these.
2309 This means three calls to sasl_listmech() without a specified mech_list.
2310 => SASL tests which mechs of all available mechs actually work
2311 => three warnings about OTP not working
2312 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2314 With the patch, sasl_listmech() also gets called three times. But now
2315 SASL's mech_list option is set to the server_mech specified in the the
2316 authenticator. Or in other words, the answer from sasl_listmech()
2317 gets limited to just the mech you're testing for (which is different
2319 => the return list contains just 'plain' or 'login', 'cram-md5' or
2320 nothing depending on the value of ob->server_mech.
2322 I've just tested the patch: Authentication still works fine,
2323 unavailable mechs specified in the exim configuration are still
2324 caught, and the auth.log warnings about OTP are gone."
2326 PH/25 When debugging is enabled, the contents of the command line are added
2327 to the debugging output, even when log_selector=+arguments is not
2330 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2331 answer is "GNU", and only if the return is "GNU/something" is the answer
2334 PH/27 $acl_verify_message is now set immediately after the failure of a
2335 verification in an ACL, and so is available in subsequent modifiers. In
2336 particular, the message can be preserved by coding like this:
2338 warn !verify = sender
2339 set acl_m0 = $acl_verify_message
2341 Previously, $acl_verify_message was set only while expanding "message"
2342 and "log_message" when a very denied access.
2344 PH/28 Modified OS/os.c-Linux with
2346 -#ifndef OS_LOAD_AVERAGE
2347 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2349 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2350 nomenclature these days.)
2352 PH/29 Installed patch from the Sieve maintainer that adds the options
2353 sieve_useraddress and sieve_subaddress to the redirect router.
2355 PH/30 In these circumstances:
2356 . Two addresses routed to the same list of hosts;
2357 . First host does not offer TLS;
2358 . First host accepts first address;
2359 . First host gives temporary error to second address;
2360 . Second host offers TLS and a TLS session is established;
2361 . Second host accepts second address.
2362 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2363 and peerdn, if requested) that were in fact used only for the second
2366 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2367 attention to any local part prefix or suffix that was matched by the
2368 router that accepted the address. It now behaves in the same way as it
2369 does for delivery: the affixes are removed from the local part unless
2370 rcpt_include_affixes is set on the transport.
2372 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2373 timeout during the DATA phase of an incoming message.
2375 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2376 applied a patch sanctioned by the Sieve maintainer.
2378 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2379 the uid or gid is negative. A case of a negative gid caused this to be
2380 noticed. The fix allows for either to be negative.
2382 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2383 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2386 PH/36 The change PH/12 above was broken. Fixed it.
2388 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2389 the grounds that routing the same address twice would always produce the
2390 same answer. This might have been true once, but it is certainly no
2391 longer true now. Routing a child address may depend on the previous
2392 routing that produced that child. Some complicated redirection strategies
2393 went wrong when messages had multiple recipients, and made Exim's
2394 behaviour dependent on the order in which the addresses were given.
2396 I have moved the duplicate checking until after the routing is complete.
2397 Exim scans the addresses that are assigned to local and remote
2398 transports, and removes any duplicates. This means that more work will be
2399 done, as duplicates will always all be routed, but duplicates are
2400 presumably rare, so I don't expect this is of any significance.
2402 For deliveries to pipes, files, and autoreplies, the duplicate checking
2403 still happens during the routing process, since they are not going to be
2406 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2407 It corrects a timeout issue with spamd. This is Ian's comment: "The
2408 background is that sometimes spamd either never reads data from a
2409 connection it has accepted, or it never writes response data. The exiscan
2410 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2411 blindly assumes that writes won't block so it may never time out."
2413 PH/39 Allow G after quota size as well as K and M.
2415 PH/40 The value set for $authenticated_id in an authenticator may not contain
2416 binary zeroes or newlines because the value is written to log lines and
2417 to spool files. There was no check on this. Now the value is run through
2418 the string_printing() function so that such characters are converted to
2419 printable escape sequences.
2421 PH/41 $message_linecount is a new variable that contains the total number of
2422 lines in the message. Compare $body_linecount, which is the count for the
2425 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2426 bounce and delay warning messages, except in certain special cases, which
2429 (a) An SMTP error message from a remote host;
2430 (b) A message specified in a :fail: redirection;
2431 (c) A message specified in a "fail" command in a system filter;
2432 (d) A message specified in a FAIL return from the queryprogram router;
2433 (e) A message specified by the cannot_route_message router option.
2435 In these cases only, Exim does include the error details in bounce and
2436 warning messages. There are also a few cases where bland messages such
2437 as "unrouteable address" or "local delivery error" are given.
2439 PH/43 $value is now also set for the "else" part of a ${run expansion.
2441 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2442 being worked on, but at least Exim now implements the latest version to
2445 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2446 process to complete was treated as a delivery failure, a timeout while
2447 writing the message to the pipe was logged, but erroneously treated as a
2448 successful delivery. Such timeouts include transport filter timeouts. For
2449 consistency with the overall process timeout, these timeouts are now
2450 treated as errors, giving rise to delivery failures by default. However,
2451 there is now a new Boolean option for the pipe transport called
2452 timeout_defer, which, if set TRUE, converts the failures into defers for
2453 both kinds of timeout. A transport filter timeout is now identified in
2456 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2457 systems where "make" and "gmake" are different, calling "gmake" at top
2458 level broke things. I've arranged for the value of $(MAKE) to be passed
2459 from the Makefile to this script so that it can call the same version of
2463 A note about Exim versions 4.44 and 4.50
2464 ----------------------------------------
2466 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2467 changes of various kinds. As a consequence, a big documentation update was
2468 needed. This delayed the release for rather longer than seemed good, especially
2469 in the light of a couple of (minor) security issues. Therefore, the changes
2470 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2471 release. So 4.44 and 4.50 are in effect two different branches that both start
2474 I have left the 4.50 change log unchanged; it contains all the changes since
2475 4.43. The change log for 4.44 is below; many of its items are identical to
2476 those for 4.50. This seems to be the most sensible way to preserve the
2477 historical information.
2483 1. Minor wording change to the doc/README.SIEVE file.
2485 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2486 computation of the current number of files was incorrect.
2488 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2489 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2492 4. Give more explanation in the error message when the command for a transport
2493 filter fails to execute.
2495 5. There are several places where Exim runs a non-Exim command in a
2496 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2497 was being done only for the command run by the queryprogram router. It is
2498 now done for all such subprocesses. The other cases are: ${run, transport
2499 filters, and the commands run by the lmtp and pipe transports.
2501 6. Added CONFIGURE_GROUP build-time option.
2503 7. Some older OS have a limit of 256 on the maximum number of file
2504 descriptors. Exim was using setrlimit() to set 1000 as a large value
2505 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2506 systems. I've change it so that if it can't get 1000, it tries for 256.
2508 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2509 was an oversight, and furthermore, ever since the addition of extra
2510 controls (e.g. 4.43/32), the checks on when to allow different forms of
2511 "control" were broken. There should now be diagnostics for all cases when a
2512 control that does not make sense is encountered.
2514 9. Added the /retain_sender option to "control=submission".
2516 10. $recipients is now available in the predata ACL (oversight).
2518 11. Tidy the search cache before the fork to do a delivery from a message
2519 received from the command line. Otherwise the child will trigger a lookup
2520 failure and thereby defer the delivery if it tries to use (for example) a
2521 cached ldap connection that the parent has called unbind on.
2523 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2524 of $address_data from the recipient verification was clobbered by the
2525 sender verification.
2527 13. The value of address_data from a sender verification is now available in
2528 $sender_address_data in subsequent conditions in the ACL statement.
2530 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2532 15. Added a new option "connect=<time>" to callout options, to set a different
2535 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2536 was its contents. (It was OK if the option was not defined at all.)
2538 17. A "Completed" log line is now written for messages that are removed from
2539 the spool by the -Mrm option.
2541 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2542 information about exactly what failed.
2544 19. Added -dd to debug only the daemon process.
2546 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2547 handles timeouts, both on the server side and network timeouts. Renamed the
2548 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2550 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2551 It is now set to "smtps".
2553 22. $host_address is now set to the target address during the checking of
2554 ignore_target_hosts.
2556 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2557 being passed; this would have caused $sender_host_name to have been used if
2558 matching the list had actually called for a host name (not very likely,
2559 since this list is usually IP addresses). A host name is now passed as
2562 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2563 code that specifies a non-restarting handler (typically sigaction() in
2564 modern systems) in an attempt to fix a rare and obscure crash bug.
2566 25. Narrowed the window for a race in the daemon that could cause it to ignore
2567 SIGCHLD signals. This is not a major problem, because they are used only to
2568 wake it up if nothing else does.
2570 26. A malformed maildirsize file could cause Exim to calculate negative values
2571 for the mailbox size or file count. Odd effects could occur as a result.
2572 The maildirsize information is now recalculated if the size or filecount
2575 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2576 support for a long time. Removed HAVE_SYS_VFS_H.
2578 28. Installed the latest version of exipick from John Jetmore.
2580 29. In an address list, if the pattern was not a regular expression, an empty
2581 subject address (from a bounce message) matched only if the pattern was an
2582 empty string. Non-empty patterns were not even tested. This was the wrong
2583 because it is perfectly reasonable to use an empty address as part of a
2584 database query. An empty address is now tested by patterns that are
2585 lookups. However, all the other forms of pattern expect the subject to
2586 contain a local part and a domain, and therefore, for them, an empty
2587 address still always fails if the pattern is not itself empty.
2589 30. Exim went into a mad DNS loop when attempting to do a callout where the
2590 host was specified on an smtp transport, and looking it up yielded more
2591 than one IP address.
2593 31. Re-factored the code for checking spool and log partition space into a
2594 function that finds that data and another that does the check. The former
2595 is then used to implement four new variables: $spool_space, $log_space,
2596 $spool_inodes, and $log_inodes.
2598 32. The RFC2047 encoding function was originally intended for short strings
2599 such as real names; it was not keeping to the 75-character limit for
2600 encoded words that the RFC imposes. It now respects the limit, and
2601 generates multiple encoded words if necessary. To be on the safe side, I
2602 have increased the buffer size for the ${rfc2047: expansion operator from
2605 33. It is now permitted to omit both strings after an "if" condition; if the
2606 condition is true, the result is "true". As before, when the second string
2607 is omitted, a false condition yields an empty string. This makes it less
2608 cumbersome to write custom ACL and router conditions.
2610 34. Failure to deliver a bounce message always caused it to be frozen, even if
2611 there was an errors_to setting on the router. The errors_to setting is now
2614 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2615 canonical form (fully expanded) before being placed in
2616 $sender_host_address.
2618 36. The table in the code that translates DNS record types into text (T_A to
2619 "A" for instance) was missing entries for NS and CNAME. It is just possible
2620 that this could have caused confusion if both these types were looked up
2621 for the same domain, because the text type is used as part of Exim's
2622 per-process caching. But the chance of anyone hitting this buglet seems
2625 37. The dnsdb lookup has been extended in a number of ways.
2627 (1) There is a new type, "zns", which walks up the domain tree until it
2628 finds some nameserver records. It should be used with care.
2630 (2) There is a new type, "mxh", which is like "mx" except that it returns
2631 just the host names, not the priorities.
2633 (3) It is now possible to give a list of domains (or IP addresses) to be
2634 looked up. The behaviour when one of the lookups defers can be
2635 controlled by a keyword.
2637 (4) It is now possible to specify the separator character for use when
2638 multiple records are returned.
2640 38. The dnslists ACL condition has been extended: it is now possible to supply
2641 a list of IP addresses and/or domains to be looked up in a particular DNS
2644 39. Added log_selector=+queue_time_overall.
2646 40. When running the queue in the test harness, wait just a tad after forking a
2647 delivery process, to get repeatability of debugging output.
2649 41. Include certificate and key file names in error message when GnuTLS fails
2650 to set them up, because the GnuTLS error message doesn't include the name
2651 of the failing file when there is a problem reading it.
2653 42. Allow both -bf and -bF in the same test run.
2655 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2657 44. The "Exiscan patch" is now merged into the mainline Exim source.
2659 45. Sometimes the final signoff response after QUIT could fail to get
2660 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2661 < 0 before doing a fflush(). This bug looks as though it goes back to the
2662 introduction of TLS in release 3.20, but "sometimes" must have been rare
2663 because the tests only now provoked it.
2665 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2666 (this can affect the format of dates).
2668 47. exim_tidydb, when checking for the continued existence of a message for
2669 which it has found a message-specific retry record, was not finding
2670 messages that were in split spool directories. Consequently, it was
2671 deleting retry records that should have stayed in existence.
2673 48. Steve fixed some bugs in eximstats.
2675 49. The SPA authentication driver was not abandoning authentication and moving
2676 on to the next authenticator when an expansion was forced to fail,
2677 contradicting the general specification for all authenticators. Instead it
2678 was generating a temporary error. It now behaves as specified.
2680 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2681 (the order specifies the preference for clients). The order is now AES256,
2682 AES128, 3DES, ARCFOUR128.
2684 51. Small patch to Sieve code - explicitly set From: when generating an
2687 52. Exim crashed if a remote delivery caused a very long error message to be
2688 recorded - for instance if somebody sent an entire SpamAssassin report back
2689 as a large number of 550 error lines. This bug was coincidentally fixed by
2690 increasing the size of one of Exim's internal buffers (big_buffer) that
2691 happened as part of the Exiscan merge. However, to be on the safe side, I
2692 have made the code more robust (and fixed the comments that describe what
2695 53. Now that there can be additional text after "Completed" in log lines (if
2696 the queue_time_overall log selector is set), a one-byte patch to exigrep
2697 was needed to allow it to recognize "Completed" as not the last thing in
2700 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2701 patch that reportedly fixes this has been added. I am not expert enough to
2702 create a test for it. This is what the patch creator wrote:
2704 "I found a little strange behaviour of ldap code when working with
2705 Windows 2003 AD Domain, where users was placed in more than one
2706 Organization Units. When I tried to give exim partial DN, the exit code
2707 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2708 But simultaneously result of request was absolutely normal ldap result,
2709 so I produce this patch..."
2711 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2712 REFERENCE, so I have modified the code to exclude the patch when that macro
2715 55. Some experimental protocols are using DNS PTR records for new purposes. The
2716 keys for these records are domain names, not reversed IP addresses. The
2717 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2718 leaves it alone. Component reversal etc. now happens only for IP addresses.
2721 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2723 57. Double the size of the debug message buffer (to 2048) so that more of very
2724 long debug lines gets shown.
2726 58. The exicyclog utility now does better if the number of log files to keep
2727 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2729 59. Two changes related to the smtp_active_hostname option:
2731 (1) $smtp_active_hostname is now available as a variable.
2732 (2) The default for smtp_banner uses $smtp_active_hostname instead
2733 of $primary_hostname.
2735 60. The host_aton() function is supposed to be passed a string that is known
2736 to be a valid IP address. However, in the case of IPv6 addresses, it was
2737 not checking this. This is a hostage to fortune. Exim now panics and dies
2738 if the condition is not met. A case was found where this could be provoked
2739 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2740 components; fortuitously, this particular loophole had already been fixed
2741 by change 4.50/55 above.
2743 If there are any other similar loopholes, the new check in host_aton()
2744 itself should stop them being exploited. The report I received stated that
2745 data on the command line could provoke the exploit when Exim was running as
2746 exim, but did not say which command line option was involved. All I could
2747 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2748 running as the user.
2751 61. There was a buffer overflow vulnerability in the SPA authentication code
2752 (which came originally from the Samba project). I have added a test to the
2753 spa_base64_to_bits() function which I hope fixes it.
2756 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2757 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2759 63. The daemon start-up calls getloadavg() while still root for those OS that
2760 need the first call to be done as root, but it missed one case: when
2761 deliver_queue_load_max is set with deliver_drop_privilege. This is
2762 necessary for the benefit of the queue runner, because there is no re-exec
2763 when deliver_drop_privilege is set.
2765 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2766 This has been fixed.
2768 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2769 was in use, was not putting the data itself into the right store pool;
2770 consequently, it could be overwritten for a subsequent message in the same
2771 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2774 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2776 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2777 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2778 the function were treating the return as a boolean value, which happened to
2779 work because 0=false and not-0=true, but is not correct code.
2781 68. The host_aton() function was not handling scoped IPv6 addresses (those
2782 with, for example, "%eth0" on the end) correctly.
2784 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2785 negated items (that is, ~something) in unsigned ints. Some compilers
2786 apparently mutter when there is no cast.
2788 70. If an address verification called from an ACL failed, and did not produce a
2789 user-specific message (i.e. there was only a "system" message), nothing was
2790 put in $acl_verify_message. In this situation, it now puts the system
2793 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2794 session; change 4.31/43 added the unwanted input to the log line - except
2795 that it did not do this in the start of session case. It now does.
2797 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2798 This is wrong and can cause the other end to generate a synchronization
2799 error if it is another Exim or anything else that does the synchronization
2800 check. A QUIT command is no longer sent after a timeout.
2802 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2803 during host lookups.
2805 74. The defer_ok option of callout verification was not working if it was used
2806 when verifying addresses in header lines, that is, for this case:
2808 verify = header_sender/callout=defer_ok
2810 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2811 those file descriptors could be used for SMTP connections. If anything
2812 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2813 could be sent to the SMTP client, causing chaos. The daemon now opens
2814 stdin, stdout, and stderr to /dev/null when it puts itself into the
2817 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2818 log by default. The user can override this with suitable Perl magic.
2820 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2821 the log message when discard triggers, was not working for the DATA ACL or
2822 for the non-SMTP ACL.
2824 78. Error message wording change in sieve.c.
2826 79. If smtp_accept_max_per_host was set, the number of connections could be
2827 restricted to fewer than expected, because the daemon was trying to set up
2828 a new connection before checking whether the processes handling previous
2829 connections had finished. The check for completed processes is now done
2830 earlier. On busy systems, this bug wouldn't be noticed because something
2831 else would have woken the daemon, and it would have reaped the completed
2834 80. If a message was submitted locally by a user whose login name contained one
2835 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2836 It caused a spool format error. I have fixed the spool reading code. A
2837 related problem was that the "from" clause in the Received: line became
2838 illegal because of the space(s). It is now covered by ${quote_local_part.
2840 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2843 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2845 83. Patch from Sieve maintainer to fix unterminated string problem in
2846 "vacation" handling.
2848 84. Some minor changes to the Linux configuration files to help with other
2849 OS variants using glibc.
2851 85. One more patch for Sieve to update vacation handling to latest spec.
2854 ----------------------------------------------------
2855 See the note above about the 4.44 and 4.50 releases.
2856 ----------------------------------------------------
2862 1. Change 4.43/35 introduced a bug that caused file counts to be
2863 incorrectly computed when quota_filecount was set in an appendfile
2866 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2867 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2870 3. Give more explanation in the error message when the command for a transport
2871 filter fails to execute.
2873 4. There are several places where Exim runs a non-Exim command in a
2874 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2875 was being done only for the command run by the queryprogram router. It is
2876 now done for all such subprocesses. The other cases are: ${run, transport
2877 filters, and the commands run by the lmtp and pipe transports.
2879 5. Some older OS have a limit of 256 on the maximum number of file
2880 descriptors. Exim was using setrlimit() to set 1000 as a large value
2881 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2882 systems. I've change it so that if it can't get 1000, it tries for 256.
2884 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2885 was an oversight, and furthermore, ever since the addition of extra
2886 controls (e.g. 4.43/32), the checks on when to allow different forms of
2887 "control" were broken. There should now be diagnostics for all cases when a
2888 control that does not make sense is encountered.
2890 7. $recipients is now available in the predata ACL (oversight).
2892 8. Tidy the search cache before the fork to do a delivery from a message
2893 received from the command line. Otherwise the child will trigger a lookup
2894 failure and thereby defer the delivery if it tries to use (for example) a
2895 cached ldap connection that the parent has called unbind on.
2897 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2898 of $address_data from the recipient verification was clobbered by the
2899 sender verification.
2901 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2902 was its contents. (It was OK if the option was not defined at all.)
2904 11. A "Completed" log line is now written for messages that are removed from
2905 the spool by the -Mrm option.
2907 12. $host_address is now set to the target address during the checking of
2908 ignore_target_hosts.
2910 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2911 being passed; this would have caused $sender_host_name to have been used if
2912 matching the list had actually called for a host name (not very likely,
2913 since this list is usually IP addresses). A host name is now passed as
2916 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2917 code that specifies a non-restarting handler (typically sigaction() in
2918 modern systems) in an attempt to fix a rare and obscure crash bug.
2920 15. Narrowed the window for a race in the daemon that could cause it to ignore
2921 SIGCHLD signals. This is not a major problem, because they are used only to
2922 wake it up if nothing else does.
2924 16. A malformed maildirsize file could cause Exim to calculate negative values
2925 for the mailbox size or file count. Odd effects could occur as a result.
2926 The maildirsize information is now recalculated if the size or filecount
2929 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2930 support for a long time. Removed HAVE_SYS_VFS_H.
2932 18. Updated exipick to current release from John Jetmore.
2934 19. Allow an empty sender to be matched against a lookup in an address list.
2935 Previously the only cases considered were a regular expression, or an
2938 20. Exim went into a mad DNS lookup loop when doing a callout where the
2939 host was specified on the transport, if the DNS lookup yielded more than
2942 21. The RFC2047 encoding function was originally intended for short strings
2943 such as real names; it was not keeping to the 75-character limit for
2944 encoded words that the RFC imposes. It now respects the limit, and
2945 generates multiple encoded words if necessary. To be on the safe side, I
2946 have increased the buffer size for the ${rfc2047: expansion operator from
2949 22. Failure to deliver a bounce message always caused it to be frozen, even if
2950 there was an errors_to setting on the router. The errors_to setting is now
2953 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2954 canonical form (fully expanded) before being placed in
2955 $sender_host_address.
2957 24. Updated eximstats to version 1.33
2959 25. Include certificate and key file names in error message when GnuTLS fails
2960 to set them up, because the GnuTLS error message doesn't include the name
2961 of the failing file when there is a problem reading it.
2963 26. Expand error message when OpenSSL has problems setting up cert/key files.
2966 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2967 (this can affect the format of dates).
2969 28. exim_tidydb, when checking for the continued existence of a message for
2970 which it has found a message-specific retry record, was not finding
2971 messages that were in split spool directories. Consequently, it was
2972 deleting retry records that should have stayed in existence.
2974 29. eximstats updated to version 1.35
2975 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2976 1.35 - bugfix such that pie charts by volume are generated correctly
2978 30. The SPA authentication driver was not abandoning authentication and moving
2979 on to the next authenticator when an expansion was forced to fail,
2980 contradicting the general specification for all authenticators. Instead it
2981 was generating a temporary error. It now behaves as specified.
2983 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2984 (the order specifies the preference for clients). The order is now AES256,
2985 AES128, 3DES, ARCFOUR128.
2987 31. Small patch to Sieve code - explicitly set From: when generating an
2990 32. Exim crashed if a remote delivery caused a very long error message to be
2991 recorded - for instance if somebody sent an entire SpamAssassin report back
2992 as a large number of 550 error lines. This bug was coincidentally fixed by
2993 increasing the size of one of Exim's internal buffers (big_buffer) that
2994 happened as part of the Exiscan merge. However, to be on the safe side, I
2995 have made the code more robust (and fixed the comments that describe what
2998 33. Some experimental protocols are using DNS PTR records for new purposes. The
2999 keys for these records are domain names, not reversed IP addresses. The
3000 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3001 leaves it alone. Component reversal etc. now happens only for IP addresses.
3004 34. The host_aton() function is supposed to be passed a string that is known
3005 to be a valid IP address. However, in the case of IPv6 addresses, it was
3006 not checking this. This is a hostage to fortune. Exim now panics and dies
3007 if the condition is not met. A case was found where this could be provoked
3008 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3009 components; fortuitously, this particular loophole had already been fixed
3010 by change 4.50/55 or 4.44/33 above.
3012 If there are any other similar loopholes, the new check in host_aton()
3013 itself should stop them being exploited. The report I received stated that
3014 data on the command line could provoke the exploit when Exim was running as
3015 exim, but did not say which command line option was involved. All I could
3016 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3017 running as the user.
3020 35. There was a buffer overflow vulnerability in the SPA authentication code
3021 (which came originally from the Samba project). I have added a test to the
3022 spa_base64_to_bits() function which I hope fixes it.
3025 36. The daemon start-up calls getloadavg() while still root for those OS that
3026 need the first call to be done as root, but it missed one case: when
3027 deliver_queue_load_max is set with deliver_drop_privilege. This is
3028 necessary for the benefit of the queue runner, because there is no re-exec
3029 when deliver_drop_privilege is set.
3031 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3032 was in use, was not putting the data itself into the right store pool;
3033 consequently, it could be overwritten for a subsequent message in the same
3034 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3037 38. Sometimes the final signoff response after QUIT could fail to get
3038 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3039 < 0 before doing a fflush(). This bug looks as though it goes back to the
3040 introduction of TLS in release 3.20, but "sometimes" must have been rare
3041 because the tests only now provoked it.
3047 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3048 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3049 Now it returns an integer. A number of calls were still expecting a T/F
3050 return. Fortuitously, in all cases, the tests worked in OK situations,
3051 which is the norm. However, things would have gone wrong on any write
3052 failures on the smtp file descriptor. This function is used when sending
3053 messages over SMTP and also when doing verify callouts.
3055 2. When Exim is called to do synchronous delivery of a locally submitted
3056 message (the -odf or -odi options), it no longer closes stderr before doing
3059 3. Implemented the mua_wrapper option.
3061 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3063 5. Implemented the functions header_remove(), header_testname(),
3064 header_add_at_position(), and receive_remove_recipient(), and exported them
3067 6. If an ACL "warn" statement specified the addition of headers, Exim already
3068 inserted X-ACL-Warn: at the start if there was no header name. However, it
3069 was not making this test for the second and subsequent header lines if
3070 there were newlines in the string. This meant that an invalid header could
3071 be inserted if Exim was badly configured.
3073 7. Allow an ACL "warn" statement to add header lines at the start or after all
3074 the Received: headers, as well as at the end.
3076 8. Added the rcpt_4xx retry error code.
3078 9. Added postmaster_mailfrom=xxx to callout verification option.
3080 10. Added mailfrom=xxxx to the callout verification option, for verify=
3083 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3084 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3086 12. Inserted some casts to stop certain compilers warning when using pointer
3087 differences as field lengths or precisions in printf-type calls (mostly
3088 affecting debugging statements).
3090 13. Added optional readline() support for -be (dynamically loaded).
3092 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3093 same clock tick as a message's arrival, so that its received time was the
3094 same as the "first fail" time on the retry record, and that message
3095 remained on the queue past the ultimate address timeout, every queue runner
3096 would try a delivery (because it was past the ultimate address timeout) but
3097 after another failure, the ultimate address timeout, which should have then
3098 bounced the address, did not kick in. This was a "< instead of <=" error;
3099 in most cases the first failure would have been in the next clock tick
3100 after the received time, and all would be well.
3102 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3103 being recognized when the domain list was tested by the match_domain
3104 condition in an expansion string.
3106 16. Added the ${str2b64: operator.
3108 17. Exim was always calling setrlimit() to set a large limit for the number of
3109 processes, without checking whether the existing limit was already
3110 adequate. (It did check for the limit on file descriptors.) Furthermore,
3111 errors from getrlimit() and setrlimit() were being ignored. Now they are
3112 logged to the main and panic logs, but Exim does carry on, to try to do its
3113 job under whatever limits there are.
3115 18. Imported PCRE 5.0.
3117 19. Trivial typo in log message " temporarily refused connection" (the leading
3120 20. If the log selector return_path_on_delivery was set and an address was
3121 redirected to /dev/null, the delivery process crashed because it assumed
3122 that a return path would always be set for a "successful" delivery. In this
3123 case, the whole delivery is bypassed as an optimization, and therefore no
3126 21. Internal re-arrangement: the function for sending a challenge and reading
3127 a response while authentication was assuming a zero-terminated challenge
3128 string. It's now changed to take a pointer and a length, to allow for
3129 binary data in such strings.
3131 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3133 23. Exim was not respecting finduser_retries when seeking the login of the
3134 uid under which it was called; it was always trying 10 times. (The default
3135 setting of finduser_retries is zero.) Also, it was sleeping after the final
3136 failure, which is pointless.
3138 24. Implemented tls_on_connect_ports.
3140 25. Implemented acl_smtp_predata.
3142 26. If the domain in control=submission is set empty, Exim assumes that the
3143 authenticated id is a complete email address when it generates From: or
3144 Sender: header lines.
3146 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3147 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3148 chown and chgrp in /bin and hostname in /usr/bin.
3150 28. Exim was keeping the "process log" file open after each use, just as it
3151 does for the main log. This opens the possibility of it remaining open for
3152 long periods when the USR1 signal hits a daemon. Occasional processlog
3153 errors were reported, that could have been caused by this. Anyway, it seems
3154 much more sensible not to leave this file open at all, so that is what now
3157 29. The long-running daemon process does not normally write to the log once it
3158 has entered its main loop, and it closes the log before doing so. This is
3159 so that log files can straightforwardly be renamed and moved. However,
3160 there are a couple of unusual error situations where the daemon does write
3161 log entries, and I had neglected to close the log afterwards.
3163 30. The text of an SMTP error response that was received during a remote
3164 delivery was being truncated at 512 bytes. This is too short for some of
3165 the long messages that one sometimes sees. I've increased the limit to
3168 31. It is now possible to make retry rules that apply only when a message has a
3169 specific sender, in particular, an empty sender.
3171 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3172 it possible to be selective about when SMTP synchronization is enforced.
3174 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3176 32. Implemented hosts_connection_nolog.
3178 33. Added an ACL for QUIT.
3180 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3183 35. Added mailbox_size and mailbox_filecount to appendfile.
3185 36. Added control = no_multiline_responses to ACLs.
3187 37. There was a bug in the logic of the code that waits for the clock to tick
3188 in the case where the clock went backwards by a substantial amount such
3189 that the microsecond fraction of "now" was more than the microsecond
3190 fraction of "then" (but the whole seconds number was less).
3192 38. Added support for the libradius Radius client library this is found on
3193 FreeBSD (previously only the radiusclient library was supported).
3199 1. When certain lookups returned multiple values in the form name=value, the
3200 quoting of the values was not always being done properly. Specifically:
3201 (a) If the value started with a double quote, but contained no whitespace,
3203 (b) If the value contained whitespace other than a space character (i.e.
3204 tabs or newlines or carriage returns) it was not quoted.
3205 This fix has been applied to the mysql and pgsql lookups by writing a
3206 separate quoting function and calling it from the lookup code. The fix
3207 should probably also be applied to nisplus, ibase and oracle lookups, but
3208 since I cannot test any of those, I have not disturbed their existing code.
3210 2. A hit in the callout cache for a specific address caused a log line with no
3211 reason for rejecting RCPT. Now it says "Previous (cached) callout
3212 verification failure".
3214 3. There was an off-by-one bug in the queryprogram router. An over-long
3215 return line was truncated at 256 instead of 255 characters, thereby
3216 overflowing its buffer with the terminating zero. As well as fixing this, I
3217 have increased the buffer size to 1024 (and made a note to document this).
3219 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3220 when Exim is waiting for an SMTP response from a remote server, Exim
3221 restarts its select() call on the socket, thereby resetting its timeout.
3222 This is not a problem when such interrupts are rare. Somebody set up a cron
3223 job to run exiwhat every 2 minutes, which is less than the normal select()
3224 timeout (5 or 10 minutes). This meant that the select() timeout never
3225 kicked in because it was always reset. I have fixed this by comparing the
3226 time when an interrupt arrives with the time at the start of the first call
3227 to select(). If more time than the timeout has elapsed, the interrupt is
3228 treated as a timeout.
3230 5. Some internal re-factoring in preparation for the addition of Sieve
3231 extensions (by MH). In particular, the "personal" test is moved to a
3232 separate function, and given an option for scanning Cc: and Bcc: (which is
3233 not set for Exim filters).
3235 6. When Exim created an email address using the login of the caller as the
3236 local part (e.g. when creating a From: or Sender: header line), it was not
3237 quoting the local part when it contained special characters such as @.
3239 7. Installed new OpenBSD configuration files.
3241 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3242 try to make them clearer.
3244 9. Callout options, other than the timeout value, were being ignored when
3245 verifying sender addresses in header lines. For example, when using
3247 verify = header_sender/callout=no_cache
3249 the cache was (incorrectly) being used.
3251 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3252 only the Cygwin environment.
3254 11. When return_path_on_delivery was set as a log selector, if different remote
3255 addresses in the same message used different return paths and parallel
3256 remote delivery occurred, the wrong values would sometimes be logged.
3257 (Whenever a remote delivery process finished, the return path value from
3258 the most recently started remote delivery process was logged.)
3260 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3261 lines when AUTH and/or TLS are in use. This is the "received protocol"
3262 field. Exim used to use "asmtp" for authenticated SMTP, without any
3263 indication (in the protocol name) for TLS use. Now it follows the RFC and
3264 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3265 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3266 names appear in log lines as well as in Received: header lines.
3268 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3269 extensions, and comparison tests, and to fix some bugs.
3271 14. Changes to the "personal" filter test:
3273 (1) The test was buggy in that it was just doing the equivalent of
3274 "contains" tests on header lines. For example, if a user's address was
3275 anne@some.where, the "personal" test would incorrectly be true for
3277 To: susanne@some.where
3279 This test is now done by extracting each address from the header in turn,
3280 and checking the entire address. Other tests that are part of "personal"
3281 are now done using regular expressions (for example, to check local parts
3282 of addresses in From: header lines).
3284 (2) The list of non-personal local parts in From: addresses has been
3285 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3286 taken from the Sieve specification recommendations.
3288 (3) If the message contains any header line starting with "List-" it is
3289 treated as non-personal.
3291 (4) The test for "circular" in the Subject: header line has been removed
3292 because it now seems ill-conceived.
3294 15. Minor typos in src/EDITME comments corrected.
3296 16. Installed latest exipick from John Jetmore.
3298 17. If headers_add on a router specified a text string that was too long for
3299 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3300 of string_sprintf() is now avoided.
3302 18. $message_body_size was not set (it was always zero) when running the DATA
3303 ACL and the local_scan() function.
3305 19. For the "mail" command in an Exim filter, no default was being set for
3306 the once_repeat time, causing a random time value to be used if "once" was
3307 specified. (If the value happened to be <= 0, no repeat happened.) The
3308 default is now 0s, meaning "never repeat". The "vacation" command was OK
3309 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3310 (I found it when inspecting the code).
3312 20. There is now an overall timeout for performing a callout verification. It
3313 defaults to 4 times the callout timeout, which applies to individual SMTP
3314 commands during the callout. The overall timeout applies when there is more
3315 than one host that can be tried. The timeout is checked before trying the
3316 next host. This prevents very long delays if there are a large number of
3317 hosts and all are timing out (e.g. when the network connections are timing
3318 out). The value of the overall timeout can be changed by specifying an
3319 additional sub-option for "callout", called "maxwait". For example:
3321 verify = sender/callout=5s,maxwait=20s
3323 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3324 to the end before writing, but this should make it even safer).
3326 22. Exim was forgetting that it had advertised PIPELINING for the second and
3327 subsequent messages on an SMTP connection. It was also not resetting its
3328 memory on STARTTLS and an internal HELO.
3330 23. When Exim logs an SMTP synchronization error within a session, it now
3331 records whether PIPELINING has been advertised or not.
3333 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3334 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3335 rather than long int.
3337 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3339 26. Added the never_mail option to autoreply.
3345 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3346 crash if the getsockname() call failed; this can happen if a connection is
3347 closed very soon after it is established. The problem was simply in the
3348 order in which certain operations were done, causing Exim to try to write
3349 to the SMTP stream before it had set up the file descriptor. The bug has
3350 been fixed by making things happen in the correct order.
3356 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3357 before the connection was closed, thus losing the rejection response.
3359 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3360 some early Solaris releases, but causes trouble in current releases where
3361 socklen_t is defined.
3363 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3366 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3369 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3370 that releasing the top of it at the end releases what was used for sub-
3371 expansions (unless the block got too big). However, discard this block if
3372 the first thing is a variable or header, so that we can use its block when
3373 it is dynamic (useful for very large $message_headers, for example).
3375 6. Lookups now cache *every* query, not just the most recent. A new, separate
3376 store pool is used for this. It can be recovered when all lookup caches are
3377 flushed. Lookups now release memory at the end of their result strings.
3378 This has involved some general refactoring of the lookup sources.
3380 7. Some code has been added to the store_xxx() functions to reduce the amount
3381 of flapping under certain conditions.
3383 8. log_incoming_interface used to affect only the <= reception log lines. Now
3384 it causes the local interface and port to be added to several more SMTP log
3385 lines, for example "SMTP connection from", and rejection lines.
3387 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3389 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3391 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3392 could be overwritten at the end of the current message (or the start of a
3393 new message if it was set in a HELO ACL). The value is now preserved for
3394 the duration of the SMTP connection.
3396 12. If a transport had a headers_rewrite setting, and a matching header line
3397 contained an unqualified address, that address was qualified, even if it
3398 did not match any rewriting rules. The underlying bug was that the values
3399 of the flags that permit the existence of unqualified sender and recipient
3400 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3401 non-local messages, and by -bnq for local messages) were not being
3402 preserved with the message after it was received.
3404 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3405 "next input=" as part of the text comprising the host identity instead of
3406 the correct text. The code was using the same buffer for two different
3407 strings. However, depending on which order the printing function evaluated
3408 its arguments, the bug did not always show up. Under Linux, for example, my
3409 test suite worked just fine.
3411 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3412 doesn't work with some older versions of Perl. It has been changed to "my",
3413 which in any case is probably the better facility to use.
3415 15. A really picky compiler found some instances of statements for creating
3416 error messages that either had too many or two few arguments for the format
3419 16. The size of the buffer for calls to the DNS resolver has been increased
3420 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3421 for addresses that have a lot of PTR records. This alleviates a problem; it
3422 does not fully solve it.
3424 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3425 buffer now truncates the list and logs the incident, which is the same
3426 action as happens when Exim is looking up a host name and its aliases.
3427 Previously in this situation something unpredictable would happen;
3428 sometimes it was "internal error: store_reset failed".
3430 18. If a server dropped the connection unexpectedly when an Exim client was
3431 using GnuTLS and trying to read a response, the client delivery process
3432 crashed while trying to generate an error log message.
3434 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3435 string, for example:
3437 warn message = H1: something\nH2: something
3439 the text was added as a single header line from Exim's point of view
3440 though it ended up OK in the delivered message. However, searching for the
3441 second and subsequent header lines using $h_h2: did not work. This has been
3442 fixed. Similarly, if a system filter added multiple headers in this way,
3443 the routers could not see them.
3445 20. Expanded the error message when iplsearch is called with an invalid key to
3446 suggest using net-iplsearch in a host list.
3448 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3449 ACLs are no longer actually imposed (and a message to that effect is
3452 22. If a "gecos" field in a passwd entry contained escaped characters, in
3453 particular, if it contained a \" sequence, Exim got it wrong when building
3454 a From: or a Sender: header from that name. A second bug also caused
3455 incorrect handling when an unquoted " was present following a character
3456 that needed quoting.
3458 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3459 was not being matched caselessly.
3461 24. Arranged for all hyphens in the exim.8 source to be escaped with
3464 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3465 properly. Recipient callout cache records were still being keyed to include
3466 the sender, even when use_sender was set false. This led to far more
3467 callouts that were necessary. The sender is no longer included in the key
3468 when use_sender is false.
3470 26. Added "control = submission" modifier to ACLs.
3472 27. Added the ${base62d: operator to decode base 62 numbers.
3474 28. dnsdb lookups can now access SRV records.
3476 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3477 the configuration file.
3479 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3480 (-v) mode. This makes the output for a verbose queue run more intelligible.
3482 31. Added a use_postmaster feature to recipient callouts.
3484 32. Added the $body_zerocount variable, containing the number of binary zero
3485 bytes in the message body.
3487 33. The time of last modification of the "new" subdirectory is now used as the
3488 "mailbox time last read" when there is a quota error for a maildir
3491 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3493 35. Added +ignore_unknown as a special item in host lists.
3495 36. Code for decoding IPv6 addresses in host lists is now included, even if
3496 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3497 address was recognized as an IP address, but was then not correctly decoded
3498 into binary, causing unexpected and incorrect effects when compared with
3505 1. Very minor rewording of debugging text in manualroute to say "list of
3506 hosts" instead of "hostlist".
3508 2. If verify=header_syntax was set, and a header line with an unqualified
3509 address (no domain) and a large number of spaces between the end of the
3510 name and the colon was received, the reception process suffered a buffer
3511 overflow, and (when I tested it) crashed. This was caused by some obsolete
3512 code that should have been removed. The fix is to remove it!
3514 3. When running in the test harness, delay a bit after writing a bounce
3515 message to get a bit more predictability in the log output.
3517 4. Added a call to search_tidyup() just before forking a reception process. In
3518 theory, someone could use a lookup in the expansion of smtp_accept_max_
3519 per_host which, without the tidyup, could leave open a database connection.
3521 5. Added the variables $recipient_data and $sender_data which get set from a
3522 lookup success in an ACL "recipients" or "senders" condition, or a router
3523 "senders" option, similar to $domain_data and $local_part_data.
3525 6. Moved the writing of debug_print from before to after the "senders" test
3528 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3529 problems for message scanning, either using a data ACL, or using
3530 local_scan() because the Received: header was not generated till after they
3531 were called (in order to set the time as the time of reception completion).
3532 I have revised the way this works. The header is now generated after the
3533 body is received, but before the ACL or local_scan() are called. After they
3534 are run, the timestamp in the header is updated.
3540 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3541 before starting a queue runner without re-exec. This happened only when
3542 deliver_drop_privilege was set or when the Exim user was set to root. The
3543 effect of the bug was that timeouts during subsequent deliveries caused
3544 crashes instead of being properly handled. The handler is now left at its
3545 default (and expected) setting.
3547 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3548 message, again when deliver_drop_privilege is set or Exim is run as root.
3549 The bug described in (1) was not present in this case, but the tidying up
3550 of the other signals was missing. I have made the two cases consistent.
3552 3. The ignore_target_hosts setting on a manualroute router was being ignored
3553 for hosts that were looked up using the /MX notation.
3555 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3558 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3559 operated on the sender address. After changing the $sender_address to <>
3560 for the sender address verify, Exim was re-instated it as the original
3561 (before rewriting) address, but remembering that it had rewritten it, so it
3562 wasn't rewriting it again. This bug also had the effect of breaking the
3563 sender address verification caching when the sender address was rewritten.
3565 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3566 This has been changed so that if the ip literal address matches
3567 ignore_target_hosts, the router declines.
3569 7. Added expansion conditions match_domain, match_address, and match_local_
3570 part (NOT match_host).
3572 8. The placeholder for the Received: header didn't have a length field set.
3574 9. Added code to Exim itself and to exim_lock to test for a specific race
3575 condition that could lead to file corruption when using MBX delivery. The
3576 issue is with the lockfile that is created in /tmp. If this file is removed
3577 after a process has opened it but before that process has acquired a lock,
3578 there is the potential for a second process to recreate the file and also
3579 acquire a lock. This could lead to two Exim processes writing to the file
3580 at the same time. The added code performs the same test as UW imapd; it
3581 checks after acquiring the lock that its file descriptor still refers to
3582 the same named file.
3584 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3585 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3588 11. Added the smtp_active_hostname option. If used, this will typically be made
3589 to depend on the incoming interface address. Because $interface_address is
3590 not set up until the daemon has forked a reception process, error responses
3591 that can happen earlier (such as "too many connections") no longer contain
3594 12. If an expansion in a condition on a "warn" statement fails because a lookup
3595 defers, the "warn" statement is abandoned, and the next ACL statement is
3596 processed. Previously this caused the whole ACL to be aborted.
3598 13. Added the iplsearch lookup type.
3600 14. Added ident_timeout as a log selector.
3602 15. Added tls_certificate_verified as a log selector.
3604 16. Added a global option tls_require_ciphers (compare the smtp transport
3605 option of the same name). This controls incoming TLS connections.
3607 17. I finally figured out how to make tls_require_ciphers do a similar thing
3608 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3609 before starting the TLS session.
3611 18. Tabs are now shown as \t in -bP output.
3613 19. If the log selector return_path_on_delivery was set, Exim crashed when
3614 bouncing a message because it had too many Received: header lines.
3616 20. If two routers both had headers_remove settings, and the first one included
3617 a superfluous trailing colon, the final name in the first list and the
3618 first name in the second list were incorrectly joined into one item (with a
3619 colon in the middle).
3625 1. Added -C and -D options to the exinext utility, mainly to make it easier
3626 to include in the automated testing, but these could be helpful when
3627 multiple configurations are in use.
3629 2. The exinext utility was not formatting the output nicely when there was
3630 an alternate port involved in the retry record key, nor when there was a
3631 message id as well (for retries that were specific to a specific message
3632 and a specific host). It was also confused by IPv6 addresses, because of
3633 the additional colons they contain. I have fixed the IPv4 problem, and
3634 patched it up to do a reasonable job for IPv6.
3636 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3637 delivery, the log line now contains "pipelined" if PIPELINING was used.
3639 4. An SMTP transport process used to panic and die if the bind() call to set
3640 an explicit outgoing interface failed. This has been changed; it is now
3641 treated in the same way as a connect() failure.
3643 5. A reference to $sender_host_name in the part of a conditional expansion
3644 that was being skipped was still causing a DNS lookup. This no longer
3647 6. The def: expansion condition was not recognizing references to header lines
3648 that used bh_ and bheader_.
3650 7. Added the _cache feature to named lists.
3652 8. The code for checking quota_filecount in the appendfile transport was
3653 allowing one more file than it should have been.
3655 9. For compatibility with Sendmail, the command line option
3663 and sets the incoming protocol and host name (for trusted callers). The
3664 host name and its colon can be omitted when only the protocol is to be set.
3665 Note the Exim already has two private options, -pd and -ps, that refer to
3666 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3667 "s", but I don't think that's a major issue.
3669 10. A number of refactoring changes to the code, none of which should affect
3672 (a) The number of logging options was getting close to filling up the
3673 32-bit word that was used as a bit map. I have split them into two classes:
3674 those that are passed in the argument to log_write(), and those that are
3675 only ever tested independently outside of that function. These are now in
3676 separate 32-bit words, so there is plenty of room for expansion again.
3677 There is no change in the user interface or the logging behaviour.
3679 (b) When building, for example, log lines, the code previously used a
3680 macro that called string_cat() twice, in order to add two strings. This is
3681 not really sufficiently general. Furthermore, there was one instance where
3682 it was actually wrong because one of the argument was used twice, and in
3683 one call a function was used. (As it happened, calling the function twice
3684 did not affect the overall behaviour.) The macro has been replaced by a
3685 function that can join an arbitrary number of extra strings onto a growing
3688 (c) The code for expansion conditions now uses a table and a binary chop
3689 instead of a serial search (which was left over from when there were very
3690 few conditions). Also, it now recognizes conditions like "pam" even when
3691 the relevant support is not compiled in: a suitably worded error message is
3692 given if an attempt is made to use such a condition.
3694 11. Added ${time_interval:xxxxx}.
3696 12. A bug was causing one of the ddress fields not to be passed back correctly
3697 from remote delivery subprocesses. The field in question was not being
3698 subsequently used, so this caused to problems in practice.
3700 13. Added new log selectors queue_time and deliver_time.
3702 14. Might have fixed a bug in maildirsizefile handling that threw up
3703 "unexpected character" debug warnings, and recalculated the data
3704 unnecessarily. In any case, I expanded the warning message to give more
3707 15. Added the message "Restricted characters in address" to the statements in
3708 the default ACL that block characters like @ and % in local parts.
3710 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3711 Three changes have been made:
3713 (a) There was a serious bug; a negative response to MAIL caused the whole
3714 recipient domain to be cached as invalid, thereby blocking all messages
3715 to all local parts at the same domain, from all senders. This bug has
3716 been fixed. The domain is no longer cached after a negative response to
3717 MAIL if the sender used is not empty.
3719 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3722 (c) A new callout option, "use_sender" has been added for people who want
3723 the modified behaviour.
3729 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3732 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3733 indeed breaks things for older releases.
3735 3. Added additional logging to the case where there is a problem reading data
3736 from a filter that is running in a subprocess using a pipe, in order to
3737 try to track down a specific problem.
3739 4. Testing facility fudge: when running in the test harness and attempting
3740 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3741 getting "No route to host". Convert this to a timeout.
3743 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3746 6. Some OS don't have socklen_t but use size_t instead. This affects the
3747 fifth argument of getsockopt() amongst other things. This is now
3748 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3749 can be set for individual OS. I have set it for SunOS5, OSF1, and
3750 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3751 some earlier ones do not.
3753 7. Change 4.30/15 was not doing the test caselessly.
3755 8. The standard form for an IPv6 address literal was being rejected by address
3756 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3757 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3758 this, as well as the form without the "IPv6" on the front (but only when
3759 address literals are enabled, of course).
3761 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3763 10. Exim crashed if a message with an empty sender address specified by -f
3764 encountered a router with an errors_to setting. This could be provoked only
3765 by a command such as
3769 where an empty string was supplied; "<>" did not hit this bug.
3771 11. Installed PCRE release 4.5.
3773 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3774 remained set. It is now erased.
3776 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3777 times from message ids (which are base 36 rather than the normal 62).
3779 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3780 were being counted as actual protocol errors, and logged if the log
3781 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3782 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3783 and DATA following a set of rejected RCPTs do not count as protocol errors.
3784 In other words, Exim assumes they were pipelined, though this may not
3785 actually be the case. Of course, in all cases the client gets an
3786 appropriate error code.
3788 15. If a lookup fails in an ACL condition, a message about the failure may
3789 be available; it is used if testing the ACL cannot continue, because most
3790 such messages specify what the cause of the deferral is. However, some
3791 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3792 that caused an old message to be retained and used if a later statement
3793 caused a defer, replacing the real cause of the deferral.
3795 16. If an IP address had so many PTR records that the DNS lookup buffer
3796 was not large enough to hold them, Exim could crash while trying to process
3797 the truncated data. It now detects and logs this case.
3799 17. Further to 4.21/58, another change has been made: if (and only if) the
3800 first line of a message (the first header line) ends with CRLF, a bare LF
3801 in a subsequent header line has a space inserted after it, so as not to
3802 terminate the header.
3804 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3805 unnecessarily, used atoi() instead of strtol(), and didn't check the
3806 termination when getting file sizes from file names by regex.
3808 19. Completely re-implemented the support for maildirsize files, in the light
3809 of a number of problems with the previous contributed implementation
3810 (4.30/29). In particular:
3812 . If the quota is zero, the maildirsize file is maintained, but no quota is
3815 . If the maildir directory does not exist, it is created before any attempt
3816 to write a maildirsize file.
3818 . The quota value in the file is just a cache; if the quota is changed in
3819 the transport, the new value overrides.
3821 . A regular expression is available for excluding directories from the
3824 20. The autoreply transport checks the characters in options that define the
3825 message's headers; it allows continued headers, but it was checking with
3826 isspace() after an embedded newline instead of explicitly looking for a
3829 21. If all the "regular" hosts to which an address was routed had passed their
3830 expiry times, and had not reached their retry times, the address was
3831 bounced, even if fallback hosts were defined. Now Exim should go on to try
3834 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3835 equivalent code in the SMTP transport. Some hosts send humungous responses
3836 to HELO/EHLO, more than 1024 it seems.
3838 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3839 gives compiler warnings in some environments. I've now done it "properly",
3842 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3843 (because of problems with the built-in one) was declared to return uschar *
3844 instead of char *, causing compiler failure.
3846 25. Fixed a file descriptor leak when processing alias/forward files.
3848 26. Fixed a minor format string issue in dbfn.c.
3850 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3852 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3853 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3856 29. When a sender address is verified, it is cached, to save repeating the test
3857 when there is more than one recipient in a message. However, when the
3858 verification involves a callout, it is possible for different callout
3859 options to be set for different recipients. It is too complicated to keep
3860 track of this in the cache, so now Exim always runs a verification when a
3861 callout is required, relying on the callout cache for the optimization.
3862 The overhead is duplication of the address routing, but this should not be
3865 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3866 to be verified with callout=postmaster, and the main callout worked but the
3867 postmaster check failed, the verification correctly failed. However, if a
3868 subsequent RCPT command asked for sender verification *without* the
3869 postmaster check, incorrect caching caused this verification also to fail,
3872 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3873 it was not caching the DNS options (qualify_single, search_parents) that
3874 were used when the lookup failed. A subsequent lookup with different
3875 options therefore always gave the same answer, though there were cases
3876 where it should not have. (Example: a "domains = !$mx_any" option on a
3877 dnslookup router: the "domains" option is always processed without any
3878 widening, but the router might have qualify_single set.) Now Exim uses the
3879 cached value only when the same options are set.
3881 32. Added John Jetmore's "exipick" utility to the distribution.
3883 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3884 than a timeout (e.g. a certificate is required, and is not provided), an
3885 Exim server now closes the connection immediately. Previously it waited for
3886 the client to close - but if the client is SSL, it seems that they each
3887 wait for each other, leading to a delay before one of them times out.
3889 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3890 maintained 0.8.x compatibility because I don't think many are using it, and
3891 it is clearly obsolete.
3893 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3896 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3897 client certificate was expired. A simple patch fixes this, though I don't
3898 understand the full logic of why the verify callback is called multiple
3901 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3902 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3903 which causes problems with some clients (such as the Certicom SSL Plus
3904 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3905 disables the coutermeasure allowing Eudora to connect."
3907 38. Exim was not checking that a write() to a log file succeeded. This could
3908 lead to Bad Things if a log got too big, in particular if it hit a file
3909 size limit. Exim now panics and dies if it cannot write to a log file, just
3910 as it does if it cannot open a log file.
3912 39. Modified OS/Makefile-Linux so that it now contains
3914 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3916 The two -D definitions ensure that Exim is compiled with large file
3917 support, which makes it possible to handle log files that are bigger than
3920 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3921 instance) a domain was checked against a named list that involved a lookup,
3922 causing $domain_data to be set, then another domain was checked against the
3923 same list, then the first domain was re-checked, the value of $domain_data
3924 after the final check could be wrong. In particular, if the second check
3925 failed, it could be set empty. This bug probably also applied to
3928 41. The strip_trailing_dot option was not being applied to the address given
3929 with the -f command-line option.
3931 42. The code for reading a message's header from the spool was incrementing
3932 $received_count, but never initializing it. This meant that the value was
3933 incorrect (doubled) while delivering a message in the same process in which
3934 it was received. In the most common configuration of Exim, this never
3935 happens - a fresh exec is done - but it can happen when
3936 deliver_drop_privilege is set.
3938 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3939 it now includes up to 150 characters of the unexpected data in the log
3942 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3943 and building data strings. The size of both of these buffers was 10 000
3944 bytes - far larger than anybody would *ever* want, thought I. Needless to
3945 say, somebody hit the limit. I have increased the maximum line length to
3946 20 000 and the maximum data length of concatenated lines to 100 000. I have
3947 also fixed two bugs, because there was no checking on these buffers. Tsk,
3948 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3949 buffer is too small.
3951 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3952 lsearch lookups. Now it does.
3954 46. When parsing a route_list item in a manualroute router, a fixed-length
3955 buffer was used for the list of hosts. I made this 1024 bytes long,
3956 thinking that nobody would ever have a list of hosts that long. Wrong.
3957 Somebody had a whole pile of complicated expansion conditions, and the
3958 string was silently truncated, leading to an expansion error. It turns out
3959 that it is easier to change to an unlimited length (owing to other changes
3960 that have happened since this code was originally written) than to build
3961 structure for giving a limitation error. The length of the item that
3962 expands into the list of hosts is now unlimited.
3964 47. The lsearch lookup could not handle data where the length of text line was
3965 more than 4095 characters. Such lines were truncated, leading to shortened
3966 data being returned. It should now handle lines of any length.
3968 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3969 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3972 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3973 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3974 no longer happens in comment lines. A list of replacements is now placed
3975 at the head of all of the source files, except those whose only change is
3976 to replace PERL_COMMAND in the very first #! line.
3978 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3979 messages on the queue, with a bottom-up merge sort, using code contributed
3980 by Michael Haardt. This should make operations like -bp somewhat faster on
3981 large queues. It won't affect queue runners, except when queue_run_in_order
3984 51. Installed eximstats 1.31 in the distribution.
3986 52. Added support for SRV lookups to the dnslookup router.
3988 53. If an ACL referred to $message_body or $message_body_end, the value was not
3989 reset for any messages that followed in the same SMTP session.
3991 54. The store-handling optimization for building very long strings was not
3992 differentiating between the different store pools. I don't think this
3993 actually made any difference in practice, but I've tidied it.
3995 55. While running the routers to verify a sender address, $sender_address
3996 was still set to the sender address. This is wrong, because when routing to
3997 send a bounce to the sender, it would be empty. Therefore, I have changed
3998 it so that, while verifying a sender address, $sender_address is set to <>.
3999 (There is no change to what happens when verifying a recipient address.)
4001 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4002 target A or AAAA records (if not already returned) without resetting the
4003 qualify_single or search_parents options of the DNS resolver. These are
4004 inappropriate in this case because the targets of MX and SRV records must
4005 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4006 target that, when qualified, matched something in the local domain. These
4007 two options are now turned off when doing these lookups.
4009 57. It seems that at least some releases of Reiserfs (which does not have the
4010 concept of a fixed number of inodes) returns zero and not -1 for the
4011 number of available inodes. This interacted badly with check_spool_inodes,
4012 which assumed that -1 was the "no such thing" setting. What I have done is
4013 to check that the total number of inodes is greater than zero before doing
4014 the test of how many are available.
4016 58. When a "warn" ACL statement has a log_message modifier, the message is
4017 remembered, and not repeated. This is to avoid a lot of repetition when a
4018 message has many recipients that cause the same warning to be written.
4019 Howewer, Exim was preserving the list of already written lines for an
4020 entire SMTP session, which doesn't seem right. The memory is now reset if a
4021 new message is started.
4023 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4024 debugging output unless log_rewrite was also set.
4026 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4027 of (int)(handle) when we know that handle contains (void *)(-1).
4029 61. The Exim daemon panic-logs an error return when it closes the incoming
4030 connection. However "connection reset by peer" seems to be common, and
4031 isn't really an error worthy of noting specially, so that particular error
4034 62. When Exim is trying to find all the local interfaces, it used to panic and
4035 die if the ioctl to get the interface flags failed. However, it seems that
4036 on at least one OS (Solaris 9) it is possible to have an interface that is
4037 included in the list of interfaces, but for which you get a failure error
4038 for this call. This happens when the interface is not "plumbed" into a
4039 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4040 failure of the "get flags" call assumes that the interface is down.
4042 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4043 makes life easier for people who are doing arithmetic on fields extracted
4044 from dates, where you often get leading zeros that should not be
4045 interpreted as octal.
4047 64. Added qualify_domain to the redirect router, to override the global
4050 65. If a pathologically long header line contained very many addresses (the
4051 report of this problem mentioned 10 000) and each of them was rewritten,
4052 Exim could use up a very large amount of memory. (It kept on making new
4053 copies of the header line as it rewrote, and never released the old ones.)
4054 At the expense of a bit more processing, the header rewriting function has
4055 been changed so that it no longer eats memory in this way.
4057 66. The generation of the Received: header has been moved from the time that a
4058 message starts to be received, to the time that it finishes. The timestamp
4059 in the Received: header should now be very close to that of the <= log
4060 line. There are two side-effects of this change:
4062 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4063 logged header lines no longer include the local Received: line, because
4064 it has not yet been created. The same applies to a copy of the message
4065 that is returned to a non-SMTP sender when a message is rejected.
4067 (b) When a filter file is tested using -bf, no additional Received: header
4068 is added to the test message. After some thought, I decided that this
4071 This change does not affect the value of $received_for. It is still set
4072 after address rewriting, but before local_scan() is called.
4074 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4076 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4077 gave an unhelpful panic error message, and a defer error. I have managed to
4078 change this behaviour so that it now rejects any supplied certificate,
4079 which seems right, as the list of acceptable certificates is empty.
4081 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4082 gave an unhelpful defer error. I have not managed to make this reject any
4083 supplied certificates, but the error message it gives is "no certificate
4084 supplied", which is not helpful.
4086 70. exigrep's output now also includes lines that are not associated with any
4087 message, but which match the given pattern. Implemented by a patch from
4088 Martin Sluka, which also tidied up the Perl a bit.
4090 71. Recipient callout verification, like sender verification, was using <> in
4091 the MAIL FROM command. This isn't really the right thing, since the actual
4092 sender may affect whether the remote host accepts the recipient or not. I
4093 have changed it to use the actual sender in the callout; this means that
4094 the cache record is now keyed on a recipient/sender pair, not just the
4095 recipient address. There doesn't seem to be a real danger of callout loops,
4096 since a callout by the remote host to check the sender would use <>.
4097 [SEE ABOVE: changed after hitting problems.]
4099 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4100 temporary errors. However, in the case of such a code being given after
4101 the end of a data transmission (i.e. after ".") Exim was failing to write
4102 a retry record for the message. (Yes, there was some broken host that was
4103 actually sending 8xx at this point.)
4105 73. An unknown lookup type in a host list could cause Exim to panic-die when
4106 the list was checked. (An example that provoked this was putting <; in the
4107 middle of a list instead of at the start.) If this happened during a DATA
4108 ACL check, a -D file could be left lying around. This kind of configuration
4109 error no longer causes Exim to die; instead it causes a defer errror. The
4110 incident is still logged to the main and panic logs.
4112 74. Buglet left over from Exim 3 conversion. The message "too many messages
4113 in one connection" was written to the rejectlog but not the mainlog, except
4114 when address rewriting (yes!) was being logged.
4116 75. Added write_rejectlog option.
4118 76. When a system filter was run not as root (that is, when system_filter_user
4119 was set), the values of the $n variables were not being returned to the
4120 main process; thus, they were not subsequently available in the $sn
4123 77. Added +return_path_on_delivery log selector.
4125 78. A connection timeout was being treated differently from recipients deferred
4126 when testing hosts_max_try with a message that was older than the host's
4127 retry timeout. (The host should not be counted, thus allowing all hosts to
4128 be tried at least once before bouncing.) This may have been the cause of an
4129 occasionally reported bug whereby a message would remain on the queue
4130 longer than the retry timeout, but would be bounced if a delivery was
4131 forced. I say "may" because I never totally pinned down the problem;
4132 setting up timeout/retry tests is difficult. See also the next item.
4134 79. The ultimate address timeout was not being applied to errors that involved
4135 a combination of host plus message (for example, a timeout on a MAIL
4136 command). When an address resolved to a number of possible hosts, and they
4137 were not all tried for each delivery (e.g. because of hosts_max_try), a
4138 message could remain on the queue longer than the retry timeout.
4140 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4143 81. Fixed an obscure SMTP outgoing bug which required at least the following
4144 conditions: (a) there was another message waiting for the same server;
4145 (b) the server returned 5xx to all RCPT commands in the first message so
4146 that the message was not completed; (c) the server dropped the connection
4147 or gave a negative response to the RSET that Exim sends to abort the
4148 transaction. The observed case was a dropped connection after DATA that had
4149 been sent in pipelining mode. That is, the server had advertised PIPELINING
4150 but was not implementing it correctly. The effect of the bug was incorrect
4151 behaviour, such as trying another host, and this could lead to a crash.
4157 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4158 and daemon.c were passed as pointers to ints; they should have been
4159 pointers to socklen_t variables (which are typically unsigned ints).
4161 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4164 3. Fixed a really odd bug that affected only the testing scheme; patching a
4165 certain fixed string in the binary changed the value of another string that
4166 happened to be identical to the end of the original first string.
4168 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4169 name", it returns that address as the IP address. On some operating
4170 systems (e.g. Solaris), it also passes back the IP address string as the
4171 "host name". However, on others (e.g. Linux), it passes back an empty
4172 string. Exim wasn't checking for this, and was changing the host name to an
4173 empty string, assuming it had been canonicized.
4175 5. Although rare, it is permitted to have more than one PTR record for a given
4176 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4177 all the names associated with an address, because they do in Solaris.
4178 However, it seems that they do not in Linux for data that comes from the
4179 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4180 I found this out when I moved to a new Linux workstation and tried to run
4181 the Exim test suite.
4183 To get round this problem I have changed the code so that it now does its
4184 own call to the DNS to look up PTR records when searching for a host name.
4185 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4186 addresses that are only in /etc/hosts are still found.
4188 This behaviour is, however, controlled by an option called host_lookup_
4189 order, which defaults to "bydns:byaddr". If people want to use the other
4190 order, or indeed, just use one or the other means of lookup, they can
4191 specify it in this variable.
4193 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4194 some operating systems, this comes back from gethostbyaddr() as an empty
4195 string, and this is what Exim used to test for. However, it seems that in
4196 other systems, "." is yielded. Exim now tests for this case too.
4198 7. The values of check_spool_space and check_log_space are now held internally
4199 as a number of kilobytes instead of an absolute number of bytes. If a
4200 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4201 kilobyte. This means that much larger values can be stored.
4203 8. Exim monitor: an attempt to get the action menu when not actually pointing
4204 at a message produces an empty menu entitled "No message selected". This
4205 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4206 no entries in it ("Shell widget menu has zero width and/or height"). So I
4207 have added a single, blank menu entry in this case.
4209 9. Added ${quote_local_part.
4211 10. MIME decoding is now applied to the contents of Subject: header lines when
4214 11. Now that a reference to $sender_host_address automatically causes a reverse
4215 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4216 host lookup before query-style lookups in lists that might use this
4217 variable. This has therefore been abolished, and the "net-" prefix is no
4218 longer necessary for query-style lookups.
4220 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4221 have been a typo for LFLAGS, so it has been changed.
4223 13. The install script calls Exim with "-C /dev/null" in order to find the
4224 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4225 to be output. Howeve, since Exim outputs its version number before the
4226 error, it didn't break the script. It just looked ugly. I fixed this by
4227 always allowing "-C /dev/null" if the caller is root.
4229 14. Ignore overlarge ACL variable number when reading spool file - insurance
4230 against a later release with more variables having written the file.
4232 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4233 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4234 this, as well as the form without the "IPv6" on the front.
4236 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4237 OS/Makefile-Darwin file.
4239 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4240 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4242 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4243 variable. However, some OS use ints for inodes. I've added cast to long int
4244 to get rid of the compiler warning.
4246 19. I had forgotten to lock out "/../" in configuration file names when
4247 ALT_CONFIG_PREFIX was set.
4249 20. Routers used for verification do not need to specify transports. However,
4250 if such a router generated a host list, and callout was configured, Exim
4251 crashed, because it could not find a port number from the (non-existent)
4252 transport. It now assumes port 25 in this circumstance.
4254 21. Added the -t option to exigrep.
4256 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4257 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4258 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4259 them is set, LOOKUP_LSEARCH is forced.
4261 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4263 24. Added sender and host information to the "rejected by local_scan()" log
4264 line; previously there was no indication of these.
4266 25. Added .include_if_exists.
4268 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4269 Linux. It turns out that not all file systems support this. Apparently some
4270 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4271 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4272 not supported on the file descriptor, is now ignored when Exim is trying to
4273 sync a directory. This applies only to Linux.
4275 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4277 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4278 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4281 29. Added support for maildirsize files from supplied patch (modified a bit).
4283 30. The use of :fail: followed by an empty string could lead Exim to respond to
4284 sender verification failures with (e.g.):
4286 550 Verification failed for <xxx>
4287 550 Sender verify failed
4289 where the first response line was missing the '-' that indicates it is not
4290 the final line of the response.
4292 31. The loop for finding the name of the user that called Exim had a hardwired
4293 limit of 10; it now uses the value of finduser_retries, which is used for
4294 all other user lookups.
4296 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4299 33. Exim was neglecting to zero errno before one call of strtol() when
4300 expanding a string and expecting an integer value. On some systems this
4301 resulted in spurious "integer overflow" errors. Also, it was casting the
4302 result into an int without checking.
4304 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4305 did not work. The code looks as if it has *never* worked, though it appears
4306 to have been documented since at least releast 1.62. I have made it work.
4308 35. The "timeout_DNS" error in retry rules, also documented since at least
4309 1.62, also never worked. As it isn't clear exactly what this means, and
4310 clearly it isn't a major issue, I have abolished the feature by treating it
4311 as "timeout", and writing a warning to the main and panic logs.
4313 36. The display of retry rules for -brt wasn't always showing the error code
4316 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4317 timeout_connect_A, timeout_connect_MX.
4319 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4320 to the empty sender.
4322 39. The daemon was not analysing the content of -oX till after it had closed
4323 stderr and disconnected from the controlling terminal. This meant that any
4324 syntax errors were only noted on the panic log, and the return code from
4325 the command was 0. By re-arranging the code a little, I've made the
4326 decoding happen first, so such errors now appear on stderr, and the return
4327 code is 1. However, the actual setting up of the sockets still happens in
4328 the disconnected process, so errors there are still only recorded on the
4331 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4332 connections (as happens on some IP stacks) was logged at start up time as
4333 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4334 it from "IPv6 and IPv4", which means that two separate sockets are being
4337 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4338 says whether AF_INET or AF_INET6 was passed as an argument.
4340 42. Exiwhat output was messed up when time zones were included in log
4343 43. Exiwhat now gives more information about the daemon's listening ports,
4344 and whether -tls-on-connect was used.
4346 44. The "port" option of the smtp transport is now expanded.
4348 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4349 silently ignored. Now an error message is written to the main and panic
4352 46. There's a new ACL modifier called "logwrite" which writes to a log file
4353 as soon as it is encountered.
4355 47. Added $local_user_uid and $local_user_gid at routing time.
4357 48. Exim crashed when trying to verify a sender address that was being
4360 49. Exim was recognizing only a space character after ".include". It now also
4361 recognizes a tab character.
4363 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4364 extracting the relevant information from the specification. The man page no
4365 longer contains scrambled data for the -d option, and I've added a section
4366 at the front about calling Exim under different names.
4368 51. Added "extra_headers" argument to the "mail" command in filter files.
4370 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4373 53. Installed eximstats 1.29.
4375 54. Added transport_filter_timeout as a generic transport option.
4377 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4378 Cc: header lines. This was required by RFC 822, but it not required by RFC
4381 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4382 incoming messages that did not have them. Now it does so only if the
4383 message originates locally, that is, if there is no associated remote host
4384 address. When Resent- header lines are present, this applies to the Resent-
4385 lines rather than the non-Resent- lines.
4387 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4388 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4390 58. Messages for configuration errors now include the name of the main
4391 configuration file - useful now that there may be more than one file in a
4392 list (.included file names were always shown).
4394 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4395 for those rare installations that do not start the daemon as root or run it
4396 setuid root. I've cut out the call to initgroups() if the daemon is not
4399 60. The Exim user and group can now be bound into the binary as text strings
4400 that are looked up at the start of Exim's processing.
4402 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4404 62. Added $mailstore_basename variable.
4406 63. Installed patch to sieve.c from Michael Haardt.
4408 64. When Exim failed to open the panic log after failing to open the main log,
4409 the original message it was trying to log was written to stderr and debug
4410 output, but if they were not available (the usual case in production), it
4411 was lost. Now it is written to syslog before the two lines that record the
4412 failures to open the logs.
4414 65. Users' Exim filters run in subprocesses under the user's uid. It is
4415 possible for a "deliver" command or an alias in a "personal" command to
4416 provoke an address rewrite. If logging of address rewriting is configured,
4417 this fails because the process is not running as root or exim. There may be
4418 a better way of dealing with this, but for the moment (because 4.30 needs
4419 to be released), I have disabled address rewrite logging when running a
4420 filter in a non-root, non-exim process.
4426 1. The buildconfig auxiliary program wasn't quoting the value set for
4427 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4428 not defined. This bug was masked in 4.22 by the effect that was fixed in
4431 2. Some messages that were rejected after a message id was allocated were
4432 shown as "incomplete" by exigrep. It no longer does this for messages that
4433 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4435 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4436 have allow_domain_literals set, the ID did not get logged in the <= line.
4437 Domain literals are now always recognized in Message-ID: header lines.
4439 4. The first argument for a ${extract expansion item is the key name or field
4440 number. Leading and trailing spaces in this item were not being ignored,
4441 causing some misleading effects.
4443 5. When deliver_drop_privilege was set, single queue runner processes started
4444 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4445 same command in the process it spins off) were not dropping privilege.
4447 6. When the daemon running as "exim" started a queue runner, it always
4448 re-executed Exim in the spun-off process. This is a waste of effort when
4449 deliver_drop_privilege is set. The new process now just calls the
4450 queue-runner function directly.
4456 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4459 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4460 ignored. Though the use of -oP was forcing the writing of a pid file, it
4461 was always written to the default place.
4463 3. If the message "no IP address found for host xxxx" is generated during
4464 incoming verification, it is now followed by identification of the incoming
4465 connection (so you can more easily find what provoked it).
4467 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4469 5. Added some features to "harden" Exim a bit more against certain attacks:
4471 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4472 be put in Local/Makefile. This is like the never_users runtime option,
4473 but it cannot be overridden. The default setting is "root".
4475 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4476 prefix string with which any file named in a -C command line option
4479 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4480 is retained for -C and -D only if the caller of Exim is root. Without
4481 it, the exim user may also use -C and -D and retain privilege.
4483 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4484 command line option is disabled.
4486 6. Macro names set by the -D option must start with an upper case letter, just
4487 like macro names defined in the configuration file.
4489 7. Added "dereference=" facility to LDAP.
4491 8. Two instances of the typo "uknown" in the source files are fixed.
4493 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4494 the Configure-Makefile script screwed up while processing it.
4496 10. Incorporated PCRE 4.4.
4498 11. The SMTP synchronization check was not operating right at the start of an
4499 SMTP session. For example, it could not catch a HELO sent before the client
4500 waited for the greeting. There is now a check for outstanding input at the
4501 point when the greeting is written. Because of the duplex, asynchronous
4502 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4503 way, but not yet received, when the check is performed.
4505 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4506 on TCP/IP sockets, because this apparently causes some broken clients to
4509 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4510 unchanged) from the Cygwin maintainer.
4512 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4513 was supported instead of testing for mbx. Effectively a typo.
4515 15. The spa authenticator server code was not checking that the input it
4516 received was valid base64.
4518 16. The debug output line for the "set" modifier in ACLs was not showing the
4519 name of the variable that was being set.
4521 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4523 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4524 lookup on its own. Something else was needed to trigger the lookup. For
4525 example, a match in host_lookup or the need for a host name in a host list.
4526 Now, if $sender_host_name is referenced and the host name has not yet been
4527 looked up, a lookup is performed. If the lookup fails, the variable remains
4528 empty, and $host_lookup_failed is set to "1".
4530 19. Added "eqi" as a case-independent comparison operator.
4532 20. The saslauthd authentication condition could segfault if neither service
4533 nor realm was specified.
4535 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4536 error message that was logged was misleading, and incoming SMTP
4537 connections were dropped. The message is now more accurate, and temporary
4538 errors are given to SMTP connections.
4540 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4541 (except RSET) with a 421 error, until QUIT is received. However, it was
4542 failing to send a response to QUIT.
4544 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4545 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4546 ACL is now run after the helo_try_verify_hosts code.
4548 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4549 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4550 case-independent, so other case variants are also recognized). Apparently
4551 some systems use these upper case variants.
4553 25. If more than two messages were waiting for the same host, and a transport
4554 filter was specified for the transport, Exim sent two messages over the
4555 same TCP/IP connection, and then failed with "socket operation on non-
4556 socket" when it tried to send the third.
4558 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4560 27. The extern definition of crypt16() in expand.c was not being excluded when
4561 the OS had its own crypt16() function.
4563 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4564 as a preferred synonym for return_size_limit, both as an option and as an
4567 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4569 30. Changed the default configuration ACL to relax the local part checking rule
4570 for addresses that are not in any local domains. For these addresses,
4571 slashes and pipe symbols are allowed within local parts, but the sequence
4572 /../ is explicitly forbidden.
4574 31. SPA server authentication was not clearing the challenge buffer before
4577 32. log_message in a "warn" ACL statement was writing to the reject log as
4578 well as to the main log, which contradicts the documentation and doesn't
4579 seem right (because no rejection is happening). So I have stopped it.
4581 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4582 However, I am unable to do any testing of this.
4584 34. Fixed an infelicity in the appendfile transport. When checking directories
4585 for a mailbox, to see if any needed to be created, it was accidentally
4586 using path names with one or more superfluous leading slashes; tracing
4587 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4589 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4590 contents are added to the log line that is written for every discarded
4591 recipient. (Previously a log_message setting was ignored.)
4593 36. The ${quote: operator now quotes the string if it is empty.
4595 37. The install script runs exim in order to find its version number. If for
4596 some reason other than non-existence or emptiness, which it checks, it
4597 could not run './exim', it was installing it with an empty version number,
4598 i.e. as "exim-". This error state is now caught, and the installation is
4601 38. An argument was missing from the function that creates an error message
4602 when Exim fails to connect to the socket for saslauthd authentication.
4603 This could cause Exim to crash, or give a corrupted message.
4605 39. Added isip, isip4, and isip6 to ${if conditions.
4607 40. The ACL variables $acl_xx are now saved with the message, and can be
4608 accessed later in routers, transports, and filters.
4610 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4611 strings in the file are not string-expanded.
4613 42. If a MAIL command specified a SIZE value that was too large to fit into an
4614 int variable, the check against message_size_limit failed. Such values are
4615 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4616 day this will have to be increased, but I don't think I want to be around
4617 when emails are that large.
4624 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4625 iconv() is not standard in FreeBSD.
4627 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4628 IPv6 enabled. The observed symptom was a segmentation fault on return from
4629 the function os_common_find_running_interfaces() in src/os.c.
4631 3. In the check_special_case() function in daemon.c I had used "errno" as an
4632 argument name, which causes warnings on some systems. This was basically a
4633 typo, since it was named "eno" in the comments!
4635 4. The code that waits for the clock to tick (at a resolution of some fraction
4636 of a second) so as to ensure message-id uniqueness was always waiting for
4637 at least one whole tick, when it could have waited for less. [This is
4638 almost certainly not relevant at current processor speeds, where it is
4639 unlikely to ever wait at all. But we try to future-proof.]
4641 5. The function that sleeps for a time interval that includes fractions of a
4642 second contained a race. It did not block SIGALRM between setting the
4643 timer, and suspending (a couple of lines later). If the interval was short
4644 and the sigsuspend() was delayed until after it had expired, the suspension
4645 never ended. On busy systems this could lead to processes getting stuck for
4648 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4649 process, before it forks any delivery processes. The open lookup caching
4650 mechanism meant that the open file or database connection was passed into
4651 the delivery process. The problem was that delivery processes always tidy
4652 up cached lookup data. This could cause a problem for the next delivery
4653 process started by the queue runner, because the external queue runner
4654 process does not know about the closure. So the next delivery process
4655 still has data in the lookup cache. In the case of a file lookup, there was
4656 no problem because closing a file descriptor in a subprocess doesn't affect
4657 the parent. However, if the lookup was caching a connection to a database,
4658 the connection was closed, and the second delivery process was likely to
4659 see errors such as "PGSQL: query failed: server closed the connection
4660 unexpectedly". The problem has been fixed by closing all cached lookups
4661 in a queue runner before running a delivery process.
4663 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4664 seem to have the "const" qualifier which it has on other OS. I've
4667 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4668 *of the same type* (client or server) with the same public name that an
4669 error should be diagnosed.
4671 9. When Exim looked up a host name for an IP address, but failed to find the
4672 original IP address when looking up the host name (a safety check), it
4673 output the message "<ip address> does not match any IP for NULL", which was
4674 confusing, to say the least. The bug was that the host name should have
4675 appeared instead of "NULL".
4677 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4678 user that is built into the binary, and the -C or -D options is used, root
4679 privilege is dropped before the configuration file is read. In addition,
4680 logging is switched to stderr instead of the normal log files. If the
4681 configuration then re-defines the Exim user, the unprivileged environment
4682 is probably not what is expected, so Exim logs a panic warning message (but
4685 However, if deliver_drop_privilege is set, the unprivileged state may well
4686 be exactly what is intended, so the warning has been cut out in that case,
4687 and Exim is allowed to try to write to its normal log files.
4693 1. smtp_return_error_details was not giving details for temporary sender
4694 or receiver verification errors.
4696 2. Diagnose a configuration error if two authenticators have the same public
4699 3. Exim used not to create the message log file for a message until the first
4700 delivery attempt. This could be confusing when incoming messages were held
4701 for policy or load reasons. The message log file is now created at the time
4702 the message is received, and an initial "Received" line is written to it.
4704 4. The automatically generated man page for command line options had a minor
4705 bug that caused no ill effects; however, a more serious problem was that
4706 the procedure for building the man page automatically didn't always
4707 operate. Consequently, release 4.20 contains an out-of-date version. This
4708 shouldn't happen again.
4710 5. When building Exim with embedded Perl support, the script that builds the
4711 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4712 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4714 6. The freeze_tell option was not being used for messages that were frozen on
4715 arrival, either by an ACL or by local_scan().
4717 7. Added the smtp_incomplete_transaction log selector.
4719 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4720 was accepting AUTH without a new EHLO.
4722 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4723 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4726 10. Logging of TCP/IP connections (when configured) now happens in the main
4727 daemon process instead of the child process, so that the TCP/IP connection
4728 count is more accurate (but it can never be perfect).
4730 11. The use of "drop" in a nested ACL was not being handled correctly in the
4731 outer ACL. Now, if condition failure induced by the nested "drop" causes
4732 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4733 or "require"), the connection is dropped.
4735 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4736 that yield "discard" can now be used with an "accept" or a "discard" verb,
4737 but an error is generated for any others (because I can't see a useful way
4738 to define what should happen).
4740 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4741 are now processed in the same way as lines in the Exim configuration file.
4742 In particular, continuation lines are supported.
4744 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4746 15. Added -ti meaning -t -i.
4748 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4749 domains, and warn by logging if others are found.
4751 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4752 returned from GIFCONF when Exim is trying to find the list of interfaces on
4753 a host. The code in os.c has been modified to copy each ifreq to an aligned
4754 structure in all cases.
4756 Also, in some cases, the returned ifreq's were being copied to a 'struct
4757 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4758 means the last couple of bytes of an IPv6 address could be chopped if the
4759 ifreq contained only a normal sockaddr (14 bytes storage).
4761 18. Named domain lists were not supported in the hosts_treat_as_local option.
4762 An entry such as +xxxx was not recognized, and was treated as a literal
4765 19. Ensure that header lines added by a DATA ACL are included in the reject log
4766 if the ACL subsequently rejects the message.
4768 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4769 MD5 (which is deprecated).
4771 21. When testing a filter file using -bf, Exim was writing a message when it
4772 took the sender from a "From " line in the message, but it was not doing so
4773 when it took $return_path from a Return-Path: header line. It now does.
4775 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4776 with a valid header line field name (a series of printing characters
4777 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4779 23. Changed "disc" in the source to "disk" to conform to the documentation and
4780 the book and for uniformity.
4782 24. Ignore Sendmail's -Ooption=value command line item.
4784 25. When execve() failed while trying to run a command in a pipe transport,
4785 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4786 could be confused with a return value of 69 from the command itself. This
4787 has been changed to 127, the value the shell returns if it is asked to run
4788 a non-existent command. The wording for the related log line suggests a
4789 non-existent command as the problem.
4791 26. If received_header_text expands to an empty string, do not add a Received:
4792 header line to the message. (Well, it adds a token one on the spool, but
4793 marks it "old" so that it doesn't get used or transmitted.)
4795 27. Installed eximstats 1.28 (addition of -nt option).
4797 28. There was no check for failure on the call to getsockname() in the daemon
4798 code. This can fail if there is a shortage of resources on the system, with
4799 ENOMEM, for example. A temporary error is now given on failure.
4801 29. Contrary to the C standard, it seems that in some environments, the
4802 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4803 program. Exim now does this explicitly; it affects the formatting of
4804 timestamps using strftime().
4806 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4807 complaints. I've now initialized all the variables, to avoid this.
4809 32. Header lines added by a system filter were not being "seen" during
4810 transport-time rewrites.
4812 33. The info_callback() function passed to OpenSSL is set up with type void
4813 (*)(SSL *, int, int), as described somewhere. However, when calling the
4814 function (actually a macro) that sets it up, the type void(*)() is
4815 expected. I've put in a cast to prevent warnings from picky compilers.
4817 34. If a DNS black list lookup found a CNAME record, but there were no A
4818 records associated with the domain it pointed at, Exim crashed.
4820 35. If a DNS black list lookup returned more than one A record, Exim ignored
4821 all but the first. It now scans all returned addresses if a particular IP
4822 value is being sought. In this situation, the contents of the
4823 $dnslist_value variable are a list of all the addresses, separated by a
4826 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4827 to accept a host name and all its aliases if the forward lookup for any of
4828 them yielded the IP address of the incoming connection. Now it accepts only
4829 those names whose forward lookup yields the correct IP address. Any other
4830 names are discarded. This closes a loophole whereby a rogue DNS
4831 administrator could create reverse DNS records to break through a
4832 wildcarded host restriction in an ACL.
4834 37. If a user filter or a system filter that ran in a subprocess used any of
4835 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4836 the wrong values were passed to the pipe command ($thisaddress had the
4837 value of $0, $0 had the value of $1, etc). This bug was introduced by
4838 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4840 38. Improved the line breaking for long SMTP error messages from ACLs.
4841 Previously, if there was no break point between 40 and 75 characters, Exim
4842 left the rest of the message alone. Two changes have been made: (a) I've
4843 reduced the minimum length to 35 characters; (b) if it can't find a break
4844 point between 35 and 75 characters, it looks ahead and uses the first one
4845 that it finds. This may give the occasional overlong line, but at least the
4846 remaining text gets split now.
4848 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4849 file descriptors might be low, and that setting 1000 would always raise it.
4850 It turns out that in some environments, the limit is already over 1000 and
4851 that lowering it causes trouble. So now Exim takes care not to decrease it.
4853 40. When delivering a message, the value of $return_path is set to $sender_
4854 address at the start of routing (routers may change the value). By an
4855 oversight, this default was not being set up when an address was tested by
4856 -bt or -bv, which affected the outcome if any router or filter referred to
4859 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4860 log only when "message" or "log_message" are set. However, if one of the
4861 conditions was an address verification, or a call to a nested ACL, the
4862 messages generated by the underlying test were being passed through. This
4863 no longer happens. The underlying message is available in $acl_verify_
4864 message for both "message" and "log_message" expansions, so it can be
4865 passed through if needed.
4867 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4868 new expansion $bh_ to give the encoded byte string without charset
4869 translation. Translation happens only if iconv() is available; HAVE_ICONV
4870 indicates this at build time. HEADERS_CHARSET gives the charset to
4871 translate to; headers_charset can change it in the configuration, and
4872 "headers charset" can change it in an individual filter file.
4874 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4875 that creates RFC 2047 encoded "words" labels them as that charset instead
4876 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4877 expansion operator; (ii) when Exim creates a From: line for a local
4878 message; (iii) when a header line is rewritten to include a "phrase" part.
4880 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4881 buggy, causing it to skip the first lines of messages whose message ID
4882 ended in 'D'. This would not have bitten before Exim release 4.14, because
4883 message IDs were unlikely to end in 'D' before then. The effect was to have
4884 incorrect size information for certain domains.
4886 45. #include "config.h" was missing at the start of the crypt16.c module. This
4887 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4890 46. If there was a timeout during a "random" callout check, Exim treated it as
4891 a failure of the random address, and carried on sending RSET and the real
4892 address. If the delay was just some slowness somewhere, the response to the
4893 original RCPT would be taken as a response to RSET and so on, causing
4894 mayhem of various kinds.
4896 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4897 when I implemented it. It didn't allow for the fact that some option values
4898 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4899 the right test for positive values.
4901 48. Domain names in DNS records are case-independent. Exim always looks them up
4902 in lower case. Some resolvers return domain names in exactly the case they
4903 appear in the zone file, that is, they may contain uppercase letters. Not
4904 all resolvers do this - some return always lower case. Exim was treating a
4905 change of case by a resolver as a change of domain, similar to a widening
4906 of a domain abbreviation. This triggered its re-routing code and so it was
4907 trying to route what was effectively the same domain again. This normally
4908 caused routing to fail (because the router wouldn't handle the domain
4909 twice). Now Exim checks for this case specially, and just changes the
4910 casing of the domain that it ultimately uses when it transmits the message
4913 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4916 50. If a filter generated a file delivery with a non-absolute name (possible if
4917 no home directory exists for the router), the forbid_file option was not
4920 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4921 the existing equality matching.
4923 52. Exim was using ints instead of ino_t variables in some places where it was
4924 dealing with inode numbers.
4926 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4927 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4928 called TMPDIR, and if it finds it is different, it changes its value.
4930 54. The smtp_printf() function is now made available to local_scan() so
4931 additional output lines can be written before returning. There is also an
4932 smtp_fflush() function to enable the detection of a dropped connection.
4933 The variables smtp_input and smtp_batched_input are exported to
4936 55. Changed the default runtime configuration: the message "Unknown user"
4937 has been removed from the ACL, and instead placed on the localuser router,
4938 using the cannot_route_message feature. This means that any verification
4939 failures that generate their own messages won't get overridden. Similarly,
4940 the "Unrouteable address" message that was in the ACL for unverifiable
4941 relay addresses has also been removed.
4943 56. Added hosts_avoid_esmtp to the smtp transport.
4945 57. The exicyclog script was not checking for the esoteric option
4946 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4947 will work only if exicyclog is run under the appropriate euid.
4949 58. Following a discussion on the list, the rules by which Exim recognises line
4950 endings on incoming messages have been changed. The -dropcr and drop_cr
4951 options are now no-ops, retained only for backwards compatibility. The
4952 following line terminators are recognized: LF CRLF CR. However, special
4953 processing applies to CR:
4955 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4956 nor a local message in the state where . is a terminator.
4958 (ii) If a bare CR is encountered in a header line, an extra space is added
4959 after the line terminator so as not to end the header. The reasoning
4960 behind this is that bare CRs in header lines are most likely either
4961 to be mistakes, or people trying to play silly games.
4963 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4964 was being incorrectly given as 18 bytes larger than it should have been.
4965 This is a VOB (very old bug).
4967 60. This may never have affected anything current, but just in case it has:
4968 When the local host is found other than at the start of a list of hosts,
4969 the local host, those with the same MX, and any that follow, are discarded.
4970 When the list in question was part of a longer list of hosts, the following
4971 hosts (not currently being processed) were also being discarded. This no
4972 longer happens. I'm not sure if this situation could ever has previously
4975 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4978 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4979 header. This is something that is recommended in a new Internet Draft, and
4980 is something that is documented as being done by Sendmail. There are two
4981 possible values. For messages generated by the autoreply transport, Exim
4984 Auto-Submitted: auto-replied
4986 whereas for all other generated messages (e.g. bounces) it adds
4988 Auto-Submitted: auto-generated
4990 63. The "personal" condition in filters now includes a test for the
4991 Auto-Submitted: header. If it contains the string "auto-" the message it
4992 not considered personal.
4994 64. Added rcpt_include_affixes as a generic transport option.
4996 65. Added queue_only_override (default true).
4998 66. Added the syslog_duplication option.
5000 67. If what should have been the first header line of a message consisted of
5001 a space followed by a colon, Exim was mis-interpreting it as a header line.
5002 It isn't of course - it is syntactically invalid and should therefore be
5003 treated as the start of the message body. The misbehaviour could have
5004 caused a number of strange effects, including loss of data in subsequent
5005 header lines, and spool format errors.
5007 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5008 client host had authenticated. This control can now be exercised by an ACL
5009 for more flexibility.
5011 69. By default, callouts do not happen when testing with -bh. There is now a
5012 variant, -bhc, which does actually run the callout code, including
5013 consulting and updating the callout cache.
5015 70. Added support for saslauthd authentication, courtesy of Alexander
5018 71. If statvfs() failed on the spool or log directories while checking their
5019 size for availability, Exim confusingly gave the error "space shortage".
5020 Furthermore, in debugging mode it crashed with a floating point exception.
5021 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5022 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5023 serious problem, Exim now writes to the main and panic logs when this
5024 happens, with details of the failure. It then refuses to accept the
5025 incoming message, giving the message "spool directory problem" or "log
5026 directory problem" with a 421 code for SMTP messages.
5028 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5029 0, 1, and 2 exist, because some OS complain for execs without them (see
5030 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5031 descriptors. However, the code omitted to check that the open succeeded,
5032 causing mysterious errors if for some reason the permissions on /dev/null
5033 got screwed. Now Exim writes a message to the main and panic logs, and
5034 bombs out if it can't open /dev/null.
5036 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5037 interact so that it is all more flexible. It is supposed to remain
5038 backwards compatible. Also added extra_local_interfaces.
5040 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5041 to bomb out with an assertion failure - to the client this appears as a
5042 connection drop. This problem occurs in the part of the code that was taken
5043 from the Samba project. Fortunately, the assertion is in a very simple
5044 function, so I have fixed this by reproducing the function inline in the
5045 one place where it is called, and arranging for authentication to fail
5046 instead of killing the process with assert().
5048 75. The SPA client code was not working when the server requested OEM rather
5049 than Unicode encoding.
5051 76. Added code to make require_files with a specific uid setting more usable in
5052 the case where statting the file as root fails - usually a non-root-mounted
5053 NFS file system. When this happens and the failure is EACCES, Exim now
5054 forks a subprocess and does the per-uid checking as the relevant uid.
5056 77. Added process_log_path.
5058 78. If log_file_path was not explicitly set, a setting of check_log_space or
5059 check_log_inodes was ignored.
5061 79. If a space check for the spool or log partitions fails, the incident is now
5062 logged. Of course, in the latter case the data may get lost...
5064 80. Added the %p formatting code to string_format() so that it can be used to
5065 print addresses in debug_print(). Adjusted all the address printing in the
5066 debugging in store.c to use %p rather than %d.
5068 81. There was a concern that a line of code in smtp_in.c could overflow a
5069 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5070 initially expressed, the concern was not well-founded, because trailing
5071 spaces are removed early. However, if the trailing spaces were followed by
5072 a NULL, they did not get removed, so the overflow was possible. Two fixes
5075 (a) I re-wrote the offending code in a cleaner fashion.
5076 (b) If an incoming SMTP command contains a NULL character, it is rejected
5079 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5080 runs initgroups(), so that if the Exim user is in any additional groups,
5081 they will be used during message reception.
5087 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob